+publisher:"University of Michigan" +contributor:("Wellman, Michael P")

University of Michigan

1. Khaliligarekani, Mohammadmahdi. Incentive Mechanisms for Managing and Controlling Cyber Risks: The Role of Cyber Insurance and Resource Pooling.

Degree: PhD, Electrical and Computer Engineering, 2020, University of Michigan

URL: http://hdl.handle.net/2027.42/155236

► Faced with a myriad of costly and frequent cyber threats, organizations not only invest in software security mechanisms such as firewalls and intrusion detection systems… (more)

▼ Faced with a myriad of costly and frequent cyber threats, organizations not only invest in software security mechanisms such as firewalls and intrusion detection systems but increasingly also turn to cyber insurance which has emerged as an accepted risk mitigation mechanism and allows purchasers of insurance policies to transfer their risks to the insurer. Insurance is fundamentally a method of risk transfer, which in general does not reduce the overall risk and may provide disincentives for firms to strengthen their security; an insured may lower its effort after purchasing coverage, a phenomenon known as moral hazard. As cyber insurance is a common method for cyber risk management, it is critical to be able to use cyber insurance as both a risk transfer mechanism and an incentive mechanism for firms to increase their security efforts. This is the central focus and main goal of this dissertation. Specifically, we consider two features of cybersecurity and their impact on the subsequent insurance contract design problem. The first is the interdependent nature of cybersecurity, whereby one entity's state of security depends not only on its own effort but also on the effort of others in the same eco-system (e.g., vendors and suppliers). The second is our ability to perform an accurate quantitative assessment of security posture at a firm-level by combining recent advances in Internet measurement and machine learning techniques. The first feature, i.e., the risk interdependence among firms is an interesting aspect that makes this contract problem different from what is typically seen in the literature: how should policies be designed for firms with dependent risk relationships? We show security interdependence leads to a profit opportunity for the insurer, created by the inefficient effort levels exerted by the insureds who do not account for risk externalities when insurance is not available. Security pre-screening then enables effective premium discrimination: firms with better security conditions may get a discount on their premium payment. This type of contract allows the insurer to take advantage of the profit opportunity by incentivizing insureds to increase their security effort and improve the state of network security. We show this conclusion holds even when an insurer has the ability to seek loss recovery when an incident can be attributed to a third party. By embedding these concepts in a practical rate-schedule based underwriting framework we show that these results can be readily implemented in existing practice. While pre-screening is an effective method to incentivize effort, the insureds may lower their efforts after the pre-screening and post-contract, within the policy period, in yet another manifestation of moral hazard. We show that this can be mitigated through periodic screening combined with premium adjustment, effectively resulting in an active policy that has built-in contingencies, and the actual premium payable is realized over time based on the screening results.… Advisors/Committee Members: Liu, Mingyan (committee member), Wellman, Michael P (committee member), Borgers, Tilman M (committee member), Naghizadeh, Prinaz (committee member), Subramanian, Vijay Gautam (committee member).

Subjects/Keywords: Cybersecurity; Cyber insurance; Contract design; Game theory; Security Economics; Optimization; Computer Science; Electrical Engineering; Industrial and Operations Engineering; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Khaliligarekani, M. (2020). Incentive Mechanisms for Managing and Controlling Cyber Risks: The Role of Cyber Insurance and Resource Pooling. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/155236

Chicago Manual of Style (16^th Edition):

Khaliligarekani, Mohammadmahdi. “Incentive Mechanisms for Managing and Controlling Cyber Risks: The Role of Cyber Insurance and Resource Pooling.” 2020. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/155236.

MLA Handbook (7^th Edition):

Khaliligarekani, Mohammadmahdi. “Incentive Mechanisms for Managing and Controlling Cyber Risks: The Role of Cyber Insurance and Resource Pooling.” 2020. Web. 27 Jan 2021.

Vancouver:

Khaliligarekani M. Incentive Mechanisms for Managing and Controlling Cyber Risks: The Role of Cyber Insurance and Resource Pooling. [Internet] [Doctoral dissertation]. University of Michigan; 2020. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/155236.

Council of Science Editors:

Khaliligarekani M. Incentive Mechanisms for Managing and Controlling Cyber Risks: The Role of Cyber Insurance and Resource Pooling. [Doctoral Dissertation]. University of Michigan; 2020. Available from: http://hdl.handle.net/2027.42/155236

2. Miehling, Erik. Reasoning Under Uncertainty in Cyber-Physical Systems: Toward Efficient and Secure Operation.

Degree: PhD, Electrical Engineering: Systems, 2018, University of Michigan

URL: http://hdl.handle.net/2027.42/144026

► The increased sensing, processing, communication, and control capabilities introduced by cyber-physical systems bring many potential improvements to the operation of society's systems, but also introduce… (more)

▼ The increased sensing, processing, communication, and control capabilities introduced by cyber-physical systems bring many potential improvements to the operation of society's systems, but also introduce questions as to how one can ensure their efficient and secure operation. This dissertation investigates three questions related to decision-making under uncertainty in cyber-physical systems settings. First, in the context of power systems and electricity markets, how can one design algorithms that guide self-interested agents to a socially optimal and physically feasible outcome, subject to the fact that agents only possess localized information of the system and can only react to local signals? The proposed algorithms, investigated in the context of two distinct models, are iterative in nature and involve the exchange of messages between agents. The first model consists of a network of interconnected power systems controlled by a collection of system operators. Each system operator possesses knowledge of its own localized region and aims to prescribe the cost minimizing set of net injections for its buses. By using relative voltage angles as messages, system operators iteratively communicate to reach a social-cost minimizing and physically feasible set of injections for the whole network. The second model consists of a market operator and market participants (distribution, generation, and transmission companies). Using locational marginal pricing, the market operator is able to guide the market participants to a competitive equilibrium, which, under an assumption on the positivity of prices, is shown to be a globally optimal solution to the non-convex social-welfare maximization problem. Common to both algorithms is the use of a quadratic power flow approximation that preserves important non-linearities (power losses) while maintaining desirable mathematical properties that permit convergence under natural conditions. Second, when a system is under attack from a malicious agent, what models are appropriate for performing real-time and scalable threat assessment and response selection when we only have partial information about the attacker's intent and capabilities? The proposed model, termed the dynamic security model, is based on a type of attack graph, termed a condition dependency graph, and describes how an attacker can infiltrate a cyber network. By embedding a state space on the graph, the model is able to quantify the attacker's progression. Consideration of multiple attacker types, corresponding to attack strategies, allows one to model the defender's uncertainty of the attacker's true strategy/intent. Using noisy security alerts, the defender maintains a belief over both the capabilities/progression of the attacker (via a security state) and its strategy (attacker type). An online, tree-based search method, termed the online defense algorithm, is developed that takes advantage of the model's structure, permitting scalable computation of defense policies. Finally, in partially observable sequential… Advisors/Committee Members: Teneketzis, Demosthenis (committee member), Wellman, Michael P (committee member), Amin, Saurabh (committee member), Cybenko, George (committee member), Mathieu, Johanna (committee member).

Subjects/Keywords: Cyber-physical systems; Security; Control theory; Partially observable Markov decision processes; Electrical Engineering; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Miehling, E. (2018). Reasoning Under Uncertainty in Cyber-Physical Systems: Toward Efficient and Secure Operation. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/144026

Chicago Manual of Style (16^th Edition):

Miehling, Erik. “Reasoning Under Uncertainty in Cyber-Physical Systems: Toward Efficient and Secure Operation.” 2018. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/144026.

MLA Handbook (7^th Edition):

Miehling, Erik. “Reasoning Under Uncertainty in Cyber-Physical Systems: Toward Efficient and Secure Operation.” 2018. Web. 27 Jan 2021.

Vancouver:

Miehling E. Reasoning Under Uncertainty in Cyber-Physical Systems: Toward Efficient and Secure Operation. [Internet] [Doctoral dissertation]. University of Michigan; 2018. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/144026.

Council of Science Editors:

Miehling E. Reasoning Under Uncertainty in Cyber-Physical Systems: Toward Efficient and Secure Operation. [Doctoral Dissertation]. University of Michigan; 2018. Available from: http://hdl.handle.net/2027.42/144026

University of Michigan

3. Wah, Elaine. Computational Models of Algorithmic Trading in Financial Markets.

Degree: PhD, Computer Science and Engineering, 2016, University of Michigan

URL: http://hdl.handle.net/2027.42/120811

► Today's trading landscape is a fragmented and complex system of interconnected electronic markets in which algorithmic traders are responsible for the majority of trading activity.… (more)

▼ Today's trading landscape is a fragmented and complex system of interconnected electronic markets in which algorithmic traders are responsible for the majority of trading activity. Questions about the effects of algorithmic trading naturally lend themselves to a computational approach, given the nature of the algorithms involved and the electronic systems in place for processing and matching orders. To better understand the economic implications of algorithmic trading, I construct computational agent-based models of scenarios with investors interacting with various algorithmic traders. I employ the simulation-based methodology of empirical game-theoretic analysis to characterize trader behavior in equilibrium under different market conditions. I evaluate the impact of algorithmic trading and market structure within three different scenarios. First, I examine the impact of a market maker on trading gains in a variety of environments. A market maker facilitates trade and supplies liquidity by simultaneously maintaining offers to buy and sell. I find that market making strongly tends to increase total welfare and the market maker is itself profitable. Market making may or may not benefit investors, however, depending on market thickness, investor impatience, and the number of trading opportunities. Second, I investigate the interplay between market fragmentation and latency arbitrage, a type of algorithmic trading strategy in which traders exercise superior speed in order to exploit price disparities between exchanges. I show that the presence of a latency arbitrageur degrades allocative efficiency in continuous markets. Periodic clearing at regular intervals, as in a frequent call market, not only eliminates the opportunity for latency arbitrage but also significantly improves welfare. Lastly, I study whether frequent call markets could potentially coexist alongside the continuous trading mechanisms employed by virtually all modern exchanges. I examine the strategic behavior of fast and slow traders who submit orders to either a frequent call market or a continuous double auction. I model this as a game of market choice, and I find strong evidence of a predator-prey relationship between fast and slow traders: the fast traders prefer to be with slower agents regardless of market, and slow traders ultimately seek the protection of the frequent call market. Advisors/Committee Members: Wellman, Michael P (committee member), Rajan, Uday (committee member), Abernethy, Jacob (committee member), Barr, Michael S (committee member).

Subjects/Keywords: algorithmic trading; high-frequency trading; frequent call market; market making; agent-based modeling; simulation; Finance; Computer Science; Business and Economics; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Wah, E. (2016). Computational Models of Algorithmic Trading in Financial Markets. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/120811

Chicago Manual of Style (16^th Edition):

Wah, Elaine. “Computational Models of Algorithmic Trading in Financial Markets.” 2016. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/120811.

MLA Handbook (7^th Edition):

Wah, Elaine. “Computational Models of Algorithmic Trading in Financial Markets.” 2016. Web. 27 Jan 2021.

Vancouver:

Wah E. Computational Models of Algorithmic Trading in Financial Markets. [Internet] [Doctoral dissertation]. University of Michigan; 2016. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/120811.

Council of Science Editors:

Wah E. Computational Models of Algorithmic Trading in Financial Markets. [Doctoral Dissertation]. University of Michigan; 2016. Available from: http://hdl.handle.net/2027.42/120811

University of Michigan

4. Cheng, Frank. Agent-Based Models for Analyzing Strategic Adaptations to Government Regulation.

Degree: PhD, Computer Science & Engineering, 2020, University of Michigan

URL: http://hdl.handle.net/2027.42/155192

► In many economic systems, participants are unable to internalize social costs. These can be anything from pollution to default risk in financial systems. To deal… (more)

▼ In many economic systems, participants are unable to internalize social costs. These can be anything from pollution to default risk in financial systems. To deal with these costs, regulators impose limits on the behavior of market participants. These regulations do not always have straightforward effects, and for new regulations a model is required to evaluate them. In this thesis I will perform this modeling task in several domains using a computational agent-based approach. This approach affords two advantages. First, agent-based models can handle intricate models of participant behavior. This is often necessary when participants are operating in a complex domain, using large modeling and computational resources of their own. Second, agent-based models combined with empirical game theoretic analysis (EGTA) can calculate Nash equilibria under new regulations. This addresses in part the Lucas critique of models with regulation, which stipulates that agents can adapt their behavior in ways that break fixed assumptions about agent behavior. I evaluate the overall effects of regulation using metrics appropriate to each domain I study. Using two models of the financial system, one based on an asset market and one based on a debt market, I study Basel regulations which have been criticized for being too simplistic and for actually being counterproductive. I find that in fact, when accounting for the strategic adaptations of banks, Basel regulations are largely beneficial for financial stability. I then examine recent EPA regulations that allow the trading of emissions credits in an attempt to bring down the cost of reducing emissions. I find that while the cost of reducing pollution is reduced as desired, costs to consumers are increased by firms that use emissions trading to coordinate price hikes. In all cases, the use of game-theoretic analysis was crucial to evaluating the effect of regulation. Advisors/Committee Members: Wellman, Michael P (committee member), Shi, Cong (committee member), Koutra, Danai (committee member), Liu, Mingyan (committee member).

Subjects/Keywords: Game theory; Economics; Networks; Banking; Basel; Emissions regulation; Computer Science; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Cheng, F. (2020). Agent-Based Models for Analyzing Strategic Adaptations to Government Regulation. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/155192

Chicago Manual of Style (16^th Edition):

Cheng, Frank. “Agent-Based Models for Analyzing Strategic Adaptations to Government Regulation.” 2020. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/155192.

MLA Handbook (7^th Edition):

Cheng, Frank. “Agent-Based Models for Analyzing Strategic Adaptations to Government Regulation.” 2020. Web. 27 Jan 2021.

Vancouver:

Cheng F. Agent-Based Models for Analyzing Strategic Adaptations to Government Regulation. [Internet] [Doctoral dissertation]. University of Michigan; 2020. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/155192.

Council of Science Editors:

Cheng F. Agent-Based Models for Analyzing Strategic Adaptations to Government Regulation. [Doctoral Dissertation]. University of Michigan; 2020. Available from: http://hdl.handle.net/2027.42/155192

University of Michigan

5. Tablante, Bartolome W. Learning and Beliefs in Non-Centralized Markets.

Degree: PhD, Economics, 2015, University of Michigan

URL: http://hdl.handle.net/2027.42/113592

► This dissertation uses microeconomic theory to examine learning and beliefs in markets that are not centralized, in order to refine our knowledge of how information… (more)

▼ This dissertation uses microeconomic theory to examine learning and beliefs in markets that are not centralized, in order to refine our knowledge of how information asymmetries affect market outcomes. I approach problems in this space from a game theoretic perspective, first building an appropriate market game to study the question of interest, then characterizing the equilibria of the game, and finally applying those characterizations to the initial problem. The dissertation consists of three distinct chapters. The first of these examines the conditions that affect learning and efficiency in a decentralized market. Past research addressing this issue has shown in several models that learning is difficult and asymmetric information is costly. In a steady state environment, the first chapter shows that both learning and efficiency are possible if and only if the good being traded is divisible. This chapter identifies the divisibility of trade as a significant friction in decentralized markets. The second chapter studies how beliefs regarding a market front-runner affect surplus in a fragmented market. A front-runner is an agent who has advance knowledge of market orders, gained from either an illegal or a technological advantage. So far, research involving front-runners has placed restrictions on the behavior of other agents, assuming that agents cannot change their strategies based on the presence of a front-runner. This work provides a framework to relax those restrictions in order to allow agents to best respond to a front-runner. The research then suggests that the impact of a front-runner on traders' strategies generally leads to a decrease in total surplus. The third chapter again contributes to the literature on learning and efficiency in a decentralized market. This chapter studies the conditions that are necessary for learning and long-run efficiency in a non-stationary environment. Past research has obtained long-run efficiency if trade is divisible and bargaining is flexible; this research shows that only divisible trade is necessary. Furthermore, while past research requires traders use complex strategies in order to learn, our research demonstrates that traders may use simple strategies, and will learn by participating in the market and observing the actions of their match partners. Advisors/Committee Members: Borgers, Tilman M. (committee member), Wellman, Michael P. (committee member), Lauermann, Stephan (committee member), Miller, David A. (committee member).

Subjects/Keywords: microeconomic theory; decentralized markets; Economics; Business and Economics

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Tablante, B. W. (2015). Learning and Beliefs in Non-Centralized Markets. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/113592

Chicago Manual of Style (16^th Edition):

Tablante, Bartolome W. “Learning and Beliefs in Non-Centralized Markets.” 2015. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/113592.

MLA Handbook (7^th Edition):

Tablante, Bartolome W. “Learning and Beliefs in Non-Centralized Markets.” 2015. Web. 27 Jan 2021.

Vancouver:

Tablante BW. Learning and Beliefs in Non-Centralized Markets. [Internet] [Doctoral dissertation]. University of Michigan; 2015. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/113592.

Council of Science Editors:

Tablante BW. Learning and Beliefs in Non-Centralized Markets. [Doctoral Dissertation]. University of Michigan; 2015. Available from: http://hdl.handle.net/2027.42/113592

University of Michigan

6. Sorg, Jonathan Daniel. The Optimal Reward Problem: Designing Effective Reward for Bounded Agents.

Degree: PhD, Computer Science & Engineering, 2011, University of Michigan

URL: http://hdl.handle.net/2027.42/89705

► In the field of reinforcement learning, agent designers build agents which seek to maximize reward. In standard practice, one reward function serves two purposes. It… (more)

▼ In the field of reinforcement learning, agent designers build agents which seek to maximize reward. In standard practice, one reward function serves two purposes. It is used to evaluate the agent and is used to directly guide agent behavior in the agent's learning algorithm. This dissertation makes four main contributions to the theory and practice of reward function design. The first is a demonstration that if an agent is bounded – if it is limited in its ability to maximize expected reward – the designer may benefit by considering two reward functions. A designer reward function is used to evaluate the agent, while a separate agent reward function is used to guide agent behavior. The designer can then solve the Optimal Reward Problem (ORP): choose the agent reward function which leads to the greatest expected reward for the designer. The second contribution is the demonstration through examples that good reward functions are chosen by assessing an agent's limitations and how they interact with the environment. An agent which maintains knowledge of the environment in the form of a Bayesian posterior distribution, but lacks adequate planning resources, can be given a reward proportional to the variance of the posterior, resulting in provably efficient exploration. An agent with poor modeling assumptions can be punished for visiting the areas of the state space it has trouble modeling, resulting in better performance. The third contribution is the Policy Gradient for Reward Design (PGRD) algorithm, a convergent gradient ascent algorithm for learning good reward functions. Experiments in multiple environments demonstrate that using PGRD for reward optimization yields better agents than using the designer's reward directly as the agent's reward. It also outperforms the use of an evaluation function at the leaf-states of the planning tree. Finally, this dissertation shows that the ORP differs from the popular work on potential-based reward shaping. Shaping rewards are constrained by properties of the environment and the designer's reward function, but they generally are defined irrespective of properties of the agent. The best shaping reward functions are suboptimal for some agents and environments. Advisors/Committee Members: Baveja, Satinder Singh (committee member), Lewis, Richard L. (committee member), Laird, John E. (committee member), Polk, Thad A. (committee member), Wellman, Michael P. (committee member).

Subjects/Keywords: Reinforcement Learning; Computer Science; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Sorg, J. D. (2011). The Optimal Reward Problem: Designing Effective Reward for Bounded Agents. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/89705

Chicago Manual of Style (16^th Edition):

Sorg, Jonathan Daniel. “The Optimal Reward Problem: Designing Effective Reward for Bounded Agents.” 2011. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/89705.

MLA Handbook (7^th Edition):

Sorg, Jonathan Daniel. “The Optimal Reward Problem: Designing Effective Reward for Bounded Agents.” 2011. Web. 27 Jan 2021.

Vancouver:

Sorg JD. The Optimal Reward Problem: Designing Effective Reward for Bounded Agents. [Internet] [Doctoral dissertation]. University of Michigan; 2011. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/89705.

Council of Science Editors:

Sorg JD. The Optimal Reward Problem: Designing Effective Reward for Bounded Agents. [Doctoral Dissertation]. University of Michigan; 2011. Available from: http://hdl.handle.net/2027.42/89705

University of Michigan

7. Hu, Junling. Learning in dynamic noncooperative multiagent systems.

Degree: PhD, Social Sciences, 1999, University of Michigan

URL: http://hdl.handle.net/2027.42/131906

► Dynamic noncooperative multiagent systems are systems where self-interested agents interact with each other and their interactions change over time. We investigate the problem of learning… (more)

▼ Dynamic noncooperative multiagent systems are systems where self-interested agents interact with each other and their interactions change over time. We investigate the problem of learning and decision making in such systems. We model the systems in the framework of general-sum stochastic games with incomplete information. We design a multiagent Q-learning method, and prove its convergence in the framework of stochastic games. The standard Q-learning method, a reinforcement learning method, was originally designed for single-agent systems. Its convergence was proved for Markov decision processes, which are single-agent problems. Our extension broadens the framework of reinforcement learning, and helps to establish the theoretical foundation for applying it to multiagent systems. We prove that our learning algorithm converges to a Nash equilibrium under certain restrictions on the game structure during learning. In our simulations of a grid-world game, the restrictions are relaxed and our learning method still converges. In addition to model-free reinforcement learning, we have also studied model-based learning where agents form models of others and update their models through observations of the environment. We find that agents' mutual learning can lead to a conjectural equilibrium, where the agents' models of the others are fulfilled, and each agent behaves optimally given its expectation. Such an equilibrium state may be suboptimal. The agents may be worse off than had they not attempted to learn the models of others at all. This poses a pitfall for multiagent learning. We also analyzed the problem of recursive modeling in a dynamic game framework. This differs from previous work which studied recursive modeling in static or repeated games. We implement various levels of recursive model in a simulated double auction market. Our experiments show that performance of an agent can be quite sensitive to its assumptions about the policies of other agents, and when there is substantial uncertainty about the level of sophistication of other agents, reducing the level of recursion might be the best policy. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Decision Making; Decision-making; Dynamic; Multiagen; Noncooperative Multiagent Systems; Q-learning

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Hu, J. (1999). Learning in dynamic noncooperative multiagent systems. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/131906

Chicago Manual of Style (16^th Edition):

Hu, Junling. “Learning in dynamic noncooperative multiagent systems.” 1999. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/131906.

MLA Handbook (7^th Edition):

Hu, Junling. “Learning in dynamic noncooperative multiagent systems.” 1999. Web. 27 Jan 2021.

Vancouver:

Hu J. Learning in dynamic noncooperative multiagent systems. [Internet] [Doctoral dissertation]. University of Michigan; 1999. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/131906.

Council of Science Editors:

Hu J. Learning in dynamic noncooperative multiagent systems. [Doctoral Dissertation]. University of Michigan; 1999. Available from: http://hdl.handle.net/2027.42/131906

University of Michigan

8. Pennock, David M. Aggregating probabilistic beliefs: Market mechanisms and graphical representations.

Degree: PhD, Statistics, 1999, University of Michigan

URL: http://hdl.handle.net/2027.42/132223

► A long-standing question in statistics is how best to aggregate the probabilistic beliefs of multiple agents. Related is the practical question of how to represent… (more)

Subjects/Keywords: Aggregating; Bayesian; Belief Aggregation; Graphical Representations; Market Mechanisms; Probabilistic Beliefs; Securities Markets

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Pennock, D. M. (1999). Aggregating probabilistic beliefs: Market mechanisms and graphical representations. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/132223

Chicago Manual of Style (16^th Edition):

Pennock, David M. “Aggregating probabilistic beliefs: Market mechanisms and graphical representations.” 1999. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/132223.

MLA Handbook (7^th Edition):

Pennock, David M. “Aggregating probabilistic beliefs: Market mechanisms and graphical representations.” 1999. Web. 27 Jan 2021.

Vancouver:

Pennock DM. Aggregating probabilistic beliefs: Market mechanisms and graphical representations. [Internet] [Doctoral dissertation]. University of Michigan; 1999. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/132223.

Council of Science Editors:

Pennock DM. Aggregating probabilistic beliefs: Market mechanisms and graphical representations. [Doctoral Dissertation]. University of Michigan; 1999. Available from: http://hdl.handle.net/2027.42/132223

University of Michigan

9. Mullen, Tracy. Design of computational market systems for network information services.

Degree: PhD, Social Sciences, 1999, University of Michigan

URL: http://hdl.handle.net/2027.42/131737

► One of the goals of an information service network is to provide users with access to a multitude of highly distributed and constantly changing information… (more)

▼ One of the goals of an information service network is to provide users with access to a multitude of highly distributed and constantly changing information sources and services. Market price systems constitute one well-studied class of mechanisms for allocating resources among distributed decision makers. By implementing a virtual market system for computational agents, we can hope to realize some of the desirable properties of markets in the distributed computing context. Moreover, the underlying economic theory provides an analytical framework for predicting aggregate behavior and for designing individual agents. This thesis identifies and explores some of the issues involved in designing computational market systems in the context of the University of Michigan Digital Library (UMDL). UMDL provides library services in a distributed environment, where information agents buy and sell information services. The explict realization of this design is to provide a commerce infrastructure that supports the process of describing, locating, and negotiating for a wide variety of information services. One part of this infrastructure is the interaction framework: service description languages, and negotiation and exchange protocols. Another consists of various infrastructure services, which simplify and automate the use of many these languages and protocols. This thesis focuses on the design of negotiation protocols and the integration of negotiation mediators, or auctions. Since the number of potential service offerings and negotiation options is unbounded, UMDL also requires some mechanism to manage the scope of markets actually available to agents. The Auction Manager infrastructure component provides commerce middleware services that simplify and automate both the creation of new markets and the matching of buyers and sellers to existing markets. Online services often allow agents to select various service options. By formally representing two selection operators (buyer and seller choice) and related inference rules, the Auction Manager can more flexibly match agents to markets and detect arbitrage opportunities. The Auction Manager also provides a vehicle to experiment with alternative auction creation policies. Two simple case studies evaluate the welfare effects of different market configurations and auction parameters, and provide a basis for future auction creation policies in the Auction Manager. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Computational Market; Design; Information Services; Library Services; Libraryservices; Network; Systems

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Mullen, T. (1999). Design of computational market systems for network information services. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/131737

Chicago Manual of Style (16^th Edition):

Mullen, Tracy. “Design of computational market systems for network information services.” 1999. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/131737.

MLA Handbook (7^th Edition):

Mullen, Tracy. “Design of computational market systems for network information services.” 1999. Web. 27 Jan 2021.

Vancouver:

Mullen T. Design of computational market systems for network information services. [Internet] [Doctoral dissertation]. University of Michigan; 1999. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/131737.

Council of Science Editors:

Mullen T. Design of computational market systems for network information services. [Doctoral Dissertation]. University of Michigan; 1999. Available from: http://hdl.handle.net/2027.42/131737

University of Michigan

10. Wurman, Peter R. Market structure and multidimensional auction design for computational economies.

Degree: PhD, Social Sciences, 1999, University of Michigan

URL: http://hdl.handle.net/2027.42/132282

► Research in multiagent systems and electronic commerce applications often involves the task of designing a mechanism to allocate scarce resources among self-interested agents. Microeconomics has… (more)

Subjects/Keywords: Computational; Design; Economies; Electronic Commerce; Market Structure; Multidimensional Auction

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Wurman, P. R. (1999). Market structure and multidimensional auction design for computational economies. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/132282

Chicago Manual of Style (16^th Edition):

Wurman, Peter R. “Market structure and multidimensional auction design for computational economies.” 1999. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/132282.

MLA Handbook (7^th Edition):

Wurman, Peter R. “Market structure and multidimensional auction design for computational economies.” 1999. Web. 27 Jan 2021.

Vancouver:

Wurman PR. Market structure and multidimensional auction design for computational economies. [Internet] [Doctoral dissertation]. University of Michigan; 1999. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/132282.

Council of Science Editors:

Wurman PR. Market structure and multidimensional auction design for computational economies. [Doctoral Dissertation]. University of Michigan; 1999. Available from: http://hdl.handle.net/2027.42/132282

University of Michigan

11. Schvartzman, Leonardo Julian. Stronger bidding strategies through empirical game-theoretic analysis and reinforcement learning.

Degree: PhD, Computer science, 2009, University of Michigan

URL: http://hdl.handle.net/2027.42/127126

► Empirical game-theoretic analysis (EGTA) combines tools from simulation, search, statistics, and game-theoretic concepts to study strategic properties of large multiagent scenarios. One direct application of… (more)

▼ Empirical game-theoretic analysis (EGTA) combines tools from simulation, search, statistics, and game-theoretic concepts to study strategic properties of large multiagent scenarios. One direct application of EGTA techniques is the study of complex market problems which, due to dynamic behavior by many agents, large strategy spaces, incomplete and incrementally revealed information, have generally resisted direct solution. I report applications of EGTA techniques to study the specific problem posed by bidding in continuous double auctions (CDAs). I develop a simulator emulating a generic CDA game commonly found in the literature, and conduct the most comprehensive CDA strategic study yet, covering versions of all major CDA strategies published to date. This EGTA study confirms prior findings about the relative performance of the different strategies. In order to improve upon existing proposals and automate the search for equilibrium strategies, I develop a general methodology to derive new strategy candidates that interleaves EGTA with reinforcement learning (RL). I apply this methodology to the CDA game, and obtain new strategies that are stronger than any other published CDA policy, culminating in a new Nash equilibrium supported by learned strategies only. I evaluate this methodology on a second scenario, the Trading Agent Competition (TAC) Travel game. Building upon an existing simulator and a dataset comprising five years of observations, I apply a similar approach to derive new CDA bidding strategies in the TAC Travel domain by interleaving EGTA and RL. New experiments confirm the superiority of the learned strategies, and find a new approximate Nash equilibrium that consists of learned strategies only. These results are evidence that the combined EGTA/RL methodology is an effective method for generating stronger strategies for CDAs, and a promising approach for other domains with similar characteristics. I formulate an iterative framework to evaluate alternative strategy exploration policies, and experimentally evaluate a set of generic policies on three market problems. I find that policies seeking beneficial deviations or best responses perform generally well, and that stochastic introduction of suboptimal deviations can often lead to more effective exploration of the strategy space in the early stages of the process. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Analysis; Bidding; Continuous Double Auctions; Empirical; Game Theory; Reinforcement Learning; Strategies; Stronger; Theoretic; Trading Agent

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Schvartzman, L. J. (2009). Stronger bidding strategies through empirical game-theoretic analysis and reinforcement learning. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/127126

Chicago Manual of Style (16^th Edition):

Schvartzman, Leonardo Julian. “Stronger bidding strategies through empirical game-theoretic analysis and reinforcement learning.” 2009. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/127126.

MLA Handbook (7^th Edition):

Schvartzman, Leonardo Julian. “Stronger bidding strategies through empirical game-theoretic analysis and reinforcement learning.” 2009. Web. 27 Jan 2021.

Vancouver:

Schvartzman LJ. Stronger bidding strategies through empirical game-theoretic analysis and reinforcement learning. [Internet] [Doctoral dissertation]. University of Michigan; 2009. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/127126.

Council of Science Editors:

Schvartzman LJ. Stronger bidding strategies through empirical game-theoretic analysis and reinforcement learning. [Doctoral Dissertation]. University of Michigan; 2009. Available from: http://hdl.handle.net/2027.42/127126

University of Michigan

12. Reeves, Daniel M. Generating trading agent strategies: Analytic and empirical methods for infinite and large games.

Degree: PhD, Computer science, 2005, University of Michigan

URL: http://hdl.handle.net/2027.42/125482

► A Strategy Generation Engine is a system that reads a description of a game or market mechanism and outputs strategies for participants. Ideally, this means… (more)

▼ A Strategy Generation Engine is a system that reads a description of a game or market mechanism and outputs strategies for participants. Ideally, this means a game solver – an algorithm to compute Nash equilibria. This is a well-studied problem and very general solutions exist, but they can only be applied to small, finite games. This thesis presents methods for finding or approximating Nash equilibria for infinite games, and for intractably large finite games. First, I define a broad class of one-shot, two-player infinite games of incomplete information. I present an algorithm for computing best-response strategies in this class and show that for many particular games the algorithm can be iterated to compute Nash equilibria. Many results from the game theory literature are reproduced – automatically – using this method, as well as novel results for new games. Next, I address the problem of finding strategies in games that, even if finite, are larger than what any exact solution method can address. Our solution involves (1) generating a small set of candidate strategies, (2) constructing via simulation an approximate payoff matrix for the simplification of the game restricted to the candidate strategies, (3) analyzing the empirical game, and (4) assessing the quality of solutions with respect to the underlying full game. I leave methods for generating candidate strategies domain-specific and focus on methods for taming the computational cost of empirical game generation. I do this by employing Monte Carlo variance reduction techniques and introducing a technique for approximating many-player games by reducing the number of players. We are additionally able to solve much larger payoff matrices than the current state-of-the-art solver by exploiting symmetry in games. I test these methods in small games with known solutions and then apply them to two realistic market scenarios: Simultaneous Ascending Auctions (SAA) and the Trading Agent Competition (TAC) travel-shopping game. For these domains I focus on two key price prediction approaches for generating candidate strategies for empirical analysis: self-confirming price predictions and Walrasian equilibrium prices. We find that these are highly effective strategies in SAA and TAC, respectively. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Analytic; Empirical; Game Theory; Generating; Infinite Games; Large; Methods; Nash Equilibrium; Strategies; Trading Agent; Trading Agents

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Reeves, D. M. (2005). Generating trading agent strategies: Analytic and empirical methods for infinite and large games. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/125482

Chicago Manual of Style (16^th Edition):

Reeves, Daniel M. “Generating trading agent strategies: Analytic and empirical methods for infinite and large games.” 2005. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/125482.

MLA Handbook (7^th Edition):

Reeves, Daniel M. “Generating trading agent strategies: Analytic and empirical methods for infinite and large games.” 2005. Web. 27 Jan 2021.

Vancouver:

Reeves DM. Generating trading agent strategies: Analytic and empirical methods for infinite and large games. [Internet] [Doctoral dissertation]. University of Michigan; 2005. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/125482.

Council of Science Editors:

Reeves DM. Generating trading agent strategies: Analytic and empirical methods for infinite and large games. [Doctoral Dissertation]. University of Michigan; 2005. Available from: http://hdl.handle.net/2027.42/125482

University of Michigan

13. Pynadath, David Varghese. Probabilistic grammars for plan recognition.

Degree: PhD, Computer science, 1999, University of Michigan

URL: http://hdl.handle.net/2027.42/131760

► A decision maker operating in the presence of a planning agent must often try to determine the plan of action driving the agent's behavior. Modeling… (more)

▼ A decision maker operating in the presence of a planning agent must often try to determine the plan of action driving the agent's behavior. Modeling the uncertainty inherent in planning domains provides a difficult challenge. If the domain representation includes an explicit probabilistic model, then the inference mechanism can compute a probability distribution over possible hypotheses, providing a sound, decision-theoretic basis for selecting optimal actions. The recognizer bases its conclusions on its uncertain a priori knowledge about the agent's mental state, its decision process, the world state, and the world's dynamics, which can be summarized by a probability distribution. It then uses its partial observations about the world to infer properties of the agent and its plan. This work first applies existing research in probabilistic context-free grammars (PCFGs) to specify this causal model and answer certain queries. This dissertation then presents new inference algorithms that generate a Bayesian network representation of the PCFG distribution. The new inference algorithms extend the set of possible queries to include posterior probability distributions over nonterminal symbols and flexible evidence handling that permits missing terminals and observations of nonterminals. However, the PCFG independence assumptions restrict the domains for which the methods are applicable. The second phase of this work defines a new model, the probabilistic state-dependent grammar (PSDG). A PSDG adds an explicit model of the external world and the agent's mental state to a PCFG model of plan selection. Production probabilities are conditioned on the values of these state variables, allowing domain specification to capture the effect of planning context on the selection process. The model also represents the world dynamics through state transition probabilities. A PSDG's explicit partition between plan and state variables facilitates both domain specification and inference, as illustrated by specifications of two example domains. The first is a driver model, constructed from scratch, that captures the planning process of a driver maneuvering on a highway. A second illustration, in the domain of air combat, demonstrates the translation of a pre-existing specification in another language into a roughly equivalent PSDG model. The PSDG model also provides practical inference algorithms. As in the PCFG case, algorithms can generate a Bayesian network representation of the underlying probability distribution. However, this work also presents specialized algorithms that exploit the particular independence properties of the PSDG language to maintain a more efficient summary of evidence in the form of a belief state. The final combination of the PSDG language model and algorithms extends the range of plan recognition domains for which practical probabilistic inference is possible. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Plan Recognition; Probabilistic Grammars

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Pynadath, D. V. (1999). Probabilistic grammars for plan recognition. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/131760

Chicago Manual of Style (16^th Edition):

Pynadath, David Varghese. “Probabilistic grammars for plan recognition.” 1999. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/131760.

MLA Handbook (7^th Edition):

Pynadath, David Varghese. “Probabilistic grammars for plan recognition.” 1999. Web. 27 Jan 2021.

Vancouver:

Pynadath DV. Probabilistic grammars for plan recognition. [Internet] [Doctoral dissertation]. University of Michigan; 1999. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/131760.

Council of Science Editors:

Pynadath DV. Probabilistic grammars for plan recognition. [Doctoral Dissertation]. University of Michigan; 1999. Available from: http://hdl.handle.net/2027.42/131760

University of Michigan

14. Liu, Chao-Lin. State-space abstraction methods for approximate evaluation of Bayesian networks.

Degree: PhD, Systems science, 1998, University of Michigan

URL: http://hdl.handle.net/2027.42/131263

► Bayesian networks provide a useful mechanism for encoding and reasoning about uncertainty. Recent progress in the design of inference algorithms for Bayesian networks has expanded… (more)

▼ Bayesian networks provide a useful mechanism for encoding and reasoning about uncertainty. Recent progress in the design of inference algorithms for Bayesian networks has expanded the acceptance of Bayesian networks into a wide range of real-world applications. Inference algorithms for Bayesian networks can return exact or approximate solutions of the probability of interest, depending on the design of the algorithms. Approximate solutions of the desired probability can be instrumental in some situations: for instance, when allocated time does not permit the computation of exact solutions and when approximate solutions already satisfy the requirements of intended applications. This thesis presents algorithms for computing approximations of conditional probabilities of interest. These algorithms employ state-space abstraction techniques for computing approximations in the forms of point-valued probabilities and bounds of probability distributions. State-space abstraction techniques aggregate states of variables for obtaining approximations at reduced computational cost. Iterative algorithms designed with these techniques have demonstrated desirable anytime performance in experiments. This thesis also discusses properties of approximations. I present theorems that relate quality of approximations to conditional independence among variables in Bayesian networks. These theorems identify variables that are affected by the state-space abstraction operations and specify how the marginal probabilities of these variables are affected. Applying these theorems, we can design control heuristics that aim to optimize the performance profiles of the iterative algorithms designed with state-space abstraction techniques. In addition, this thesis reports conditions under which we can apply the state-space abstraction techniques to compute bounds of distributions. These bounds tighten over computation time, and can be useful for a variety of applications. In particular, these bounds of probability distribution can be applied to resolve tradeoffs in qualitative probabilistic networks. This thesis discusses the applications of state-space abstraction and incremental node marginalization techniques to this tradeoff resolution problem. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Approximate; Bayesian; Evaluation; Methods; Networks; Node Marginalization; State-space Abstraction

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Liu, C. (1998). State-space abstraction methods for approximate evaluation of Bayesian networks. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/131263

Chicago Manual of Style (16^th Edition):

Liu, Chao-Lin. “State-space abstraction methods for approximate evaluation of Bayesian networks.” 1998. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/131263.

MLA Handbook (7^th Edition):

Liu, Chao-Lin. “State-space abstraction methods for approximate evaluation of Bayesian networks.” 1998. Web. 27 Jan 2021.

Vancouver:

Liu C. State-space abstraction methods for approximate evaluation of Bayesian networks. [Internet] [Doctoral dissertation]. University of Michigan; 1998. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/131263.

Council of Science Editors:

Liu C. State-space abstraction methods for approximate evaluation of Bayesian networks. [Doctoral Dissertation]. University of Michigan; 1998. Available from: http://hdl.handle.net/2027.42/131263

University of Michigan

15. Walsh, William Edward. Market protocols for decentralized supply chain formation.

Degree: PhD, Computer science, 2001, University of Michigan

URL: http://hdl.handle.net/2027.42/126745

► In order to effectively respond to changing market conditions, business partners must be able to rapidly form supply chains. This thesis approaches the problem of… (more)

▼ In order to effectively respond to changing market conditions, business partners must be able to rapidly form supply chains. This thesis approaches the problem of automating supply chain formation – the process of determining the participants in a supply chain, who will exchange what with whom, and the terms of the exchanges – within an economic framework. In this thesis, supply chain formation is formalized as task dependency networks. This model captures subtask decomposition in the presence of resource contention – two important and challenging aspects of supply chain formation. In order to form supply chains in a decentralized fashion, price systems provide an economic framework for guiding the decisions of self-interested agents. In competitive price equilibrium, agents choose optimal allocations with respect to prices, and outcomes are optimal overall. Approximate competitive equilibria yield approximately optimal allocations. Different market protocols are proposed for agents to negotiate the allocation of resources to form supply chains. In the presence of resource contention, these protocols produce better solutions than the greedy protocols common in the artificial intelligence and multiagent systems literature. The first protocol proposed is based on distributed, progressive, price-based auctions, and is analyzed with non-strategic, agent bidding policies. The protocol often converges to high-value supply chains, and when competitive equilibria exist, typically to approximate competitive equilibria. However, complementarities in agent production technologies can cause the protocol to wastefully allocate inputs to agents that do not produce their outputs. A subsequent decommitment phase recovers a significant fraction of the lost surplus. Alternatively, the problem can be solved with a combinatorial auction protocol, which ensures that agents do not receive partial allocations. A combinatorial protocol is analyzed with various strategic bidding policies, and empirically compared with the distributed, price-based protocol according to various economic metrics. The distributed, price-based protocol can also be used to solve propositional satisfiability problems. Although the performance of the most direct implementation is slow, variants perform well compared with non-market algorithms. Advisors/Committee Members: Wellman, Michael P. (advisor).

Subjects/Keywords: Decentralized Supply Chain; Distributed Constraint Satisfaction; Market Protocols; Multiagent Systems; Supply Chain Formation

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Walsh, W. E. (2001). Market protocols for decentralized supply chain formation. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/126745

Chicago Manual of Style (16^th Edition):

Walsh, William Edward. “Market protocols for decentralized supply chain formation.” 2001. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/126745.

MLA Handbook (7^th Edition):

Walsh, William Edward. “Market protocols for decentralized supply chain formation.” 2001. Web. 27 Jan 2021.

Vancouver:

Walsh WE. Market protocols for decentralized supply chain formation. [Internet] [Doctoral dissertation]. University of Michigan; 2001. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/126745.

Council of Science Editors:

Walsh WE. Market protocols for decentralized supply chain formation. [Doctoral Dissertation]. University of Michigan; 2001. Available from: http://hdl.handle.net/2027.42/126745

16. Kutty, Sindhu Krishnan. All Men Count with You, but None Too Much: Information Aggregation and Learning in Prediction Markets.

Degree: PhD, Computer Science and Engineering, 2015, University of Michigan

URL: http://hdl.handle.net/2027.42/111398

► Prediction markets are markets that are set up to aggregate information from a population of traders in order to predict the outcome of an event.… (more)

Subjects/Keywords: Prediction Markets; Machine Learning; Recommender Systems; Scoring Rules; Computer Science; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Kutty, S. K. (2015). All Men Count with You, but None Too Much: Information Aggregation and Learning in Prediction Markets. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/111398

Chicago Manual of Style (16^th Edition):

Kutty, Sindhu Krishnan. “All Men Count with You, but None Too Much: Information Aggregation and Learning in Prediction Markets.” 2015. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/111398.

MLA Handbook (7^th Edition):

Kutty, Sindhu Krishnan. “All Men Count with You, but None Too Much: Information Aggregation and Learning in Prediction Markets.” 2015. Web. 27 Jan 2021.

Vancouver:

Kutty SK. All Men Count with You, but None Too Much: Information Aggregation and Learning in Prediction Markets. [Internet] [Doctoral dissertation]. University of Michigan; 2015. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/111398.

Council of Science Editors:

Kutty SK. All Men Count with You, but None Too Much: Information Aggregation and Learning in Prediction Markets. [Doctoral Dissertation]. University of Michigan; 2015. Available from: http://hdl.handle.net/2027.42/111398

17. Wolfe, Britton D. Modeling Dynamical Systems with Structured Predictive State Representations.

Degree: PhD, Computer Science & Engineering, 2009, University of Michigan

URL: http://hdl.handle.net/2027.42/64601

► Predictive state representations (PSRs) are a class of models that represent the state of a dynamical system as a set of predictions about future events.… (more)

▼ Predictive state representations (PSRs) are a class of models that represent the state of a dynamical system as a set of predictions about future events. PSRs can model partially observable, stochastic dynamical systems, including any system that can be modeled by a finite partially observable Markov decision process (POMDP). Using PSR models can help an artificial intelligence agent learn an accurate model of its environment (which is a dynamical system) from its experience in that environment. Specifically, I present the suffix-history algorithm and demonstrate that it can learn PSR models that are generally more accurate than POMDP models learned from the same amount of experience. The suffix-history algorithm learns a type of PSR called the linear PSR. However, it is intractable to learn a linear PSR (or a POMDP) to model large systems because these models do not take advantage of regularities or structure in the environment. Therefore, I present three new classes of PSR models that exploit different types of structure in an environment: hierarchical PSRs, factored PSRs, and multi-mode PSRs. Hierarchical PSRs exploit temporal structure in the environment, because a temporally abstract model can be simpler than a fully-detailed model. I demonstrate that learning a hierarchical PSR is tractable in environments in which learning a single linear PSR is intractable. Factored PSRs model systems with vector-valued observations, exploiting conditional independence among the components of the observation vectors. Leveraging that conditional independence can lead to a factored PSR model that is exponentially smaller than an unstructured model of the same system. Finally, multi-mode PSRs model systems that switch among several modes of operation. The modes used by multi-mode PSRs are defined in terms of past and future observations, which leads to advantages both when learning the model and when using it to make predictions. For each class of structured PSR models, I develop a learning algorithm that scales to larger systems than the suffix-history algorithm but still leverages the advantage of predictive state for learning accurate models. Advisors/Committee Members: Baveja, Satinder Singh (committee member), Durfee, Edmund H. (committee member), Eustice, Ryan M. (committee member), Wellman, Michael P. (committee member).

Subjects/Keywords: Predictive State Representations; Artificial Intelligence; Modeling Dynamical Systems; Computer Science; Engineering

7 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Wolfe, B. D. (2009). Modeling Dynamical Systems with Structured Predictive State Representations. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/64601

Chicago Manual of Style (16^th Edition):

Wolfe, Britton D. “Modeling Dynamical Systems with Structured Predictive State Representations.” 2009. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/64601.

MLA Handbook (7^th Edition):

Wolfe, Britton D. “Modeling Dynamical Systems with Structured Predictive State Representations.” 2009. Web. 27 Jan 2021.

Vancouver:

Wolfe BD. Modeling Dynamical Systems with Structured Predictive State Representations. [Internet] [Doctoral dissertation]. University of Michigan; 2009. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/64601.

Council of Science Editors:

Wolfe BD. Modeling Dynamical Systems with Structured Predictive State Representations. [Doctoral Dissertation]. University of Michigan; 2009. Available from: http://hdl.handle.net/2027.42/64601

18. Duong, Quang A. Graphical Multiagent Models.

Degree: PhD, Computer Science and Engineering, 2012, University of Michigan

URL: http://hdl.handle.net/2027.42/95999

► I introduce Graphical Multiagent Models (GMMs): probabilistic graphical models that capture agent interactions in factored representations for efficient inference about agent behaviors. The graphical model… (more)

▼ I introduce Graphical Multiagent Models (GMMs): probabilistic graphical models that capture agent interactions in factored representations for efficient inference about agent behaviors. The graphical model formalism exploits locality of agent interactions to support compact expression, and provides a repertoire of inference algorithms for efficient reasoning about joint behavior. I demonstrate that the GMM representation is sufficiently flexible to accommodate diverse sources of knowledge about agent behavior, and to combine these for improved predictions. History-dependent GMMs (hGMMs) extend the static form of GMMs to support representation and reasoning about multiagent behavior over time, offering the particular advantage of capturing joint dependencies induced by abstraction of history information. Computational experiments demonstrate the benefits of explicitly modeling joint behavior and allowing expression of action dependencies when given limited history information, compared to alternative models. Many multiagent modeling tasks that employ probabilistic models entail constructing dependence graph structures from observational data. In the context of graphical games where agents' payoffs depend only on actions of agents in their local neighborhoods, I formally describe the problem of learning payoff dependence structures from limited payoff observations, and investigate several learning algorithms based on minimizing empirical loss. I show that similar algorithms can also be applied to learning both dependence structures and parameters for hGMMs from time-series data. I employ data on human-subject experiments in constructing hGMMs and evaluating the learned models' prediction performance for a consensus dynamics scenario. Analysis of learned graphical structures reveals patterns of action dependence not directly reflected in observed interactions. The problem of modeling information diffusion on partially observed networks provides another demonstration of hGMM flexibility, as unobserved edges may induce correlations among node states. As graphical multiagent models can compensate for correlations induced from missing edges, I find that learning graphical multiagent models for a given network structure from diffusion traces can outperform directly recovering the missing edges, across various network settings. Advisors/Committee Members: Wellman, Michael P. (committee member), Nguyen, Long (committee member), Baveja, Satinder Singh (committee member), Durfee, Edmund H. (committee member).

Subjects/Keywords: Multiagent Reasoning; Graphical Models; Dynamics Behavior; Structure Learning; Computer Science; Engineering; Science

6 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Duong, Q. A. (2012). Graphical Multiagent Models. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/95999

Chicago Manual of Style (16^th Edition):

Duong, Quang A. “Graphical Multiagent Models.” 2012. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/95999.

MLA Handbook (7^th Edition):

Duong, Quang A. “Graphical Multiagent Models.” 2012. Web. 27 Jan 2021.

Vancouver:

Duong QA. Graphical Multiagent Models. [Internet] [Doctoral dissertation]. University of Michigan; 2012. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/95999.

Council of Science Editors:

Duong QA. Graphical Multiagent Models. [Doctoral Dissertation]. University of Michigan; 2012. Available from: http://hdl.handle.net/2027.42/95999

19. Boerkoel Jr, James C. Distributed Approaches for Solving Constraint-based Multiagent Scheduling Problems.

Degree: PhD, Computer Science & Engineering, 2012, University of Michigan

URL: http://hdl.handle.net/2027.42/96046

► This research focuses on building foundational algorithms for scheduling agents that assist people in managing their activities in environments in which tempo and complexity outstrip… (more)

▼ This research focuses on building foundational algorithms for scheduling agents that assist people in managing their activities in environments in which tempo and complexity outstrip people's cognitive capacity. The critical challenge is that, as individuals decide how to act on their scheduling goals, scheduling agents should answer queries regarding the events in their interacting schedules while respecting individual privacy and autonomy to the extent possible. I formally define both the Multiagent Simple Temporal Problem (MaSTP) and Multiagent Disjunctive Temporal Problem (MaDTP) for naturally capturing and reasoning over the distributed but interconnected scheduling problems of multiple individuals. My hypothesis is that combining a bottom-up approach – where an agent externalizes constraints that compactly summarize how its local subproblem affects other agents' subproblems, with a top-down approach – where an agent proactively constructs and internalizes new local constraints that decouple its subproblem from others', will lead to effective solution techniques. I confirm that my hypothesized approach leads to distributed algorithms that calculate summaries of the joint solution space for multiagent scheduling problems, without centralizing or otherwise redistributing the problems. In both the MaSTP and MaDTP domains, the distributed algorithms permit concurrent execution for significant speedup over current art, and also increase the level of privacy and independence in individual agent reasoning. These algorithms are most advantageous for problems where interactions between the agents are sparse compared to the complexity of agents' individual scheduling problems. Moreover, despite the combinatorially-large and unwieldy nature of the MaDTP solution space, I show that agents can use influence spaces, which compactly capture the impact of agents' interacting schedules, to tractably converge on distributed summaries of the joint solution space. Finally, I apply the same basic principle to the Hybrid Scheduling Problem, which combines constraint-based scheduling with a rudimentary level of planning, and show that my Hybrid Constraint Tightening precompilation algorithm can improve the propagation of information between planning and scheduling subproblems, leading to significant search space pruning and execution time reduction. Advisors/Committee Members: Durfee, Edmund H. (committee member), Cohn, Amy Ellen (committee member), Pollack, Martha E. (committee member), Wellman, Michael P. (committee member).

Subjects/Keywords: Multiagent Scheduling; Constraint-based Scheduling; Simple Temporal Problem; Disjunctive Temporal Problem; Temporal Decoupling; Multiagent Coordination; Computer Science; Engineering

10 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Boerkoel Jr, J. C. (2012). Distributed Approaches for Solving Constraint-based Multiagent Scheduling Problems. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/96046

Chicago Manual of Style (16^th Edition):

Boerkoel Jr, James C. “Distributed Approaches for Solving Constraint-based Multiagent Scheduling Problems.” 2012. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/96046.

MLA Handbook (7^th Edition):

Boerkoel Jr, James C. “Distributed Approaches for Solving Constraint-based Multiagent Scheduling Problems.” 2012. Web. 27 Jan 2021.

Vancouver:

Boerkoel Jr JC. Distributed Approaches for Solving Constraint-based Multiagent Scheduling Problems. [Internet] [Doctoral dissertation]. University of Michigan; 2012. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/96046.

Council of Science Editors:

Boerkoel Jr JC. Distributed Approaches for Solving Constraint-based Multiagent Scheduling Problems. [Doctoral Dissertation]. University of Michigan; 2012. Available from: http://hdl.handle.net/2027.42/96046

20. Cohn, Robert W. Maximizing Expected Value of Information in Decision Problems by Querying on a Wish-to-Know Basis.

Degree: PhD, Computer Science and Engineering, 2016, University of Michigan

URL: http://hdl.handle.net/2027.42/120772

► An agent acting under uncertainty regarding how it should complete the task assigned to it by its human user can learn more about how it… (more)

▼ An agent acting under uncertainty regarding how it should complete the task assigned to it by its human user can learn more about how it should behave by posing queries to its human user. Asking too many queries, however, may risk requiring undue attentional demand of the user, and so the agent should prioritize asking the most valuable queries. For decision-making agents, Expected Value of Information (EVOI) measures the value of a query, and so given a set of queries the agent can ask, the agent should ask the query that is expected to maximally improve its performance by selecting the query with highest EVOI in its set. Unfortunately, to compute the EVOI of a query, the agent must consider how each possible response would influence its future behavior, which makes query selection particularly challenging in settings where planning the agent's behavior would be expensive even without the added complication of considering queries to ask, especially when there are many potential queries the agent should consider. The focus of this dissertation is on developing query selection algorithms that can be feasibly applied to such settings. The main novel approach studied, Wishful Query Projection (WQP), is based on the intuition that the agent should consider which query to ask on the basis of obtaining knowledge that would help it resolve a particular dilemma that it wishes could be resolved, as opposed to blindly searching its entire query set in hopes of finding one that would give it valuable knowledge. In implementing WQP, this dissertation contributes algorithms that are founded upon the following novel result: for myopic settings, when the agent can ask any query as long as the query has no more than some set number of possible responses, the best query takes the form of asking the user to choose from a specified subset of ways for the agent to behave. The work presented shows that WQP selects queries with near-optimal EVOI when the agent's query set is (1) balanced in the range of queries it contains; and (2) rich in terms of the highest contained query EVOI. Advisors/Committee Members: Baveja, Satinder Singh (committee member), Durfee, Edmund H (committee member), Lewis, Richard L (committee member), Wellman, Michael P (committee member).

Subjects/Keywords: Expected Value of Information; Query Selection Algorithms; Computer Science; Engineering

5 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Cohn, R. W. (2016). Maximizing Expected Value of Information in Decision Problems by Querying on a Wish-to-Know Basis. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/120772

Chicago Manual of Style (16^th Edition):

Cohn, Robert W. “Maximizing Expected Value of Information in Decision Problems by Querying on a Wish-to-Know Basis.” 2016. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/120772.

MLA Handbook (7^th Edition):

Cohn, Robert W. “Maximizing Expected Value of Information in Decision Problems by Querying on a Wish-to-Know Basis.” 2016. Web. 27 Jan 2021.

Vancouver:

Cohn RW. Maximizing Expected Value of Information in Decision Problems by Querying on a Wish-to-Know Basis. [Internet] [Doctoral dissertation]. University of Michigan; 2016. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/120772.

Council of Science Editors:

Cohn RW. Maximizing Expected Value of Information in Decision Problems by Querying on a Wish-to-Know Basis. [Doctoral Dissertation]. University of Michigan; 2016. Available from: http://hdl.handle.net/2027.42/120772

21. Vorobeychik, Yevgeniy. Mechanism Design and Analysis Using Simulation-Based Game Models.

Degree: PhD, Computer Science & Engineering, 2008, University of Michigan

URL: http://hdl.handle.net/2027.42/60786

► As agent technology matures, it becomes easier to envision electronic marketplaces teeming with autonomous agents. Since agents are explicitly programmed to (nearly) optimally compete in… (more)

▼ As agent technology matures, it becomes easier to envision electronic marketplaces teeming with autonomous agents. Since agents are explicitly programmed to (nearly) optimally compete in these marketplaces, and markets themselves are designed with specific objectives in mind, tools are necessary for systematic analyses of strategic interactions among autonomous agents. While traditional game-theoretic approaches to the analysis of multi-agent systems can provide much insight, they are often inadequate, as they rely heavily on analytic tractability of the problem at hand; however, even mildly realistic models of electronic marketplaces contain enough complexity to render a fully analytic approach hopeless. To address questions not amenable to traditional theoretical approaches, I develop methods that allow systematic computational analysis of game-theoretic models in which the players' payoff functions are represented using simulations (i.e., simulation-based games). I develop a globally convergent algorithm for Nash equilibrium approximation in infinite simulation-based games, which I instantiate in the context of infinite games of incomplete information. Additionally, I use statistical learning techniques to improve the quality of Nash equilibrium approximation based on data collected from a game simulator. I also derive probabilistic confidence bounds and present convergence results about solutions of finite games modeled using simulations. The former allow an analyst to make statistically-founded statements about results based on game-theoretic simulations, while the latter provide formal justification for approximating game-theoretic solutions using simulation experiments. To address the broader mechanism design problem, I introduce an iterative algorithm for search in the design space, which requires a game solver as a subroutine. As a result, I enable computational mechanism design using simulation-based models of games by availing the designer of a set of solution tools geared specifically towards games modeled using simulations. I apply the developed computational techniques to analyze strategic procurement and answer design questions in a supply-chain simulation, as well as to analyze dynamic bidding strategies in sponsored search auctions. Indeed, the techniques I develop have broad potential applicability beyond electronic marketplaces: they are geared towards any system that features competing strategic players who respond to incentives in a way that can be reasonably predicted via a game-theoretic analysis. Advisors/Committee Members: Wellman, Michael P. (committee member), Baveja, Satinder Singh (committee member), Durfee, Edmund H. (committee member), Ozdenoren, Emre (committee member).

Subjects/Keywords: Simulation-based Game Theory; Computational Game Theory; Multiagent Systems; Computer Science; Engineering

14 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Vorobeychik, Y. (2008). Mechanism Design and Analysis Using Simulation-Based Game Models. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/60786

Chicago Manual of Style (16^th Edition):

Vorobeychik, Yevgeniy. “Mechanism Design and Analysis Using Simulation-Based Game Models.” 2008. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/60786.

MLA Handbook (7^th Edition):

Vorobeychik, Yevgeniy. “Mechanism Design and Analysis Using Simulation-Based Game Models.” 2008. Web. 27 Jan 2021.

Vancouver:

Vorobeychik Y. Mechanism Design and Analysis Using Simulation-Based Game Models. [Internet] [Doctoral dissertation]. University of Michigan; 2008. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/60786.

Council of Science Editors:

Vorobeychik Y. Mechanism Design and Analysis Using Simulation-Based Game Models. [Doctoral Dissertation]. University of Michigan; 2008. Available from: http://hdl.handle.net/2027.42/60786

22. Wiedenbeck, Bryce. Approximate Analysis of Large Simulation-Based Games.

Degree: PhD, Computer Science and Engineering, 2015, University of Michigan

URL: http://hdl.handle.net/2027.42/113587

► Game theory offers powerful tools for reasoning about agent behavior and incentives in multi-agent systems. Traditional approaches to game-theoretic analysis require enumeration of all possible… (more)

▼ Game theory offers powerful tools for reasoning about agent behavior and incentives in multi-agent systems. Traditional approaches to game-theoretic analysis require enumeration of all possible strategies and outcomes. This often constrains game models to small numbers of agents and strategies or simple closed-form payoff descriptions. Simulation-based game theory extends the reach of game-theoretic analysis through the use of agent-based modeling. In the simulation-based approach, the analyst describes an environment procedurally and then computes payoffs by simulation of agent interactions in that environment. I use simulation-based game theory to study a model of credit network formation. Credit networks represent trust relationships in a directed graph and have been proposed as a mechanism for distributed transactions without a central currency. I explore what information is important when agents make initial decisions of whom to trust, and what sorts of networks can result from their decisions. This setting demonstrates both the value of simulation-based game theory—extending game-theoretic analysis beyond analytically tractable models—and its limitations—simulations produce prodigious amounts of data, and the number of simulations grows exponentially in the number of agents and strategies. I propose several techniques for approximate analysis of simulation-based games with large numbers of agents and large amounts of simulation data. First, I show how bootstrap-based statistics can be used to estimate confidence bounds on the results of simulation-based game analysis. I show that bootstrap confidence intervals for regret of approximate equilibria are well-calibrated. Next, I describe deviation-preserving reduction, which approximates an environment with a large number of agents using a game model with a small number of players, and demonstrate that it outperforms previous player reductions on several measures. Finally, I employ machine learning to construct game models from sparse data sets, and provide evidence that learned game models can produce even better approximate equilibria in large games than deviation-preserving reduction. Advisors/Committee Members: Wellman, Michael P. (committee member), Borgers, Tilman M. (committee member), Baveja, Satinder Singh (committee member), Schoenebeck, Grant (committee member).

Subjects/Keywords: simulation-based game theory; Computer Science; Engineering

12 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Wiedenbeck, B. (2015). Approximate Analysis of Large Simulation-Based Games. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/113587

Chicago Manual of Style (16^th Edition):

Wiedenbeck, Bryce. “Approximate Analysis of Large Simulation-Based Games.” 2015. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/113587.

MLA Handbook (7^th Edition):

Wiedenbeck, Bryce. “Approximate Analysis of Large Simulation-Based Games.” 2015. Web. 27 Jan 2021.

Vancouver:

Wiedenbeck B. Approximate Analysis of Large Simulation-Based Games. [Internet] [Doctoral dissertation]. University of Michigan; 2015. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/113587.

Council of Science Editors:

Wiedenbeck B. Approximate Analysis of Large Simulation-Based Games. [Doctoral Dissertation]. University of Michigan; 2015. Available from: http://hdl.handle.net/2027.42/113587

23. Cassell, Ben-Alexander. Scaling Empirical Game-Theoretic Analysis.

Degree: PhD, Computer Science and Engineering, 2014, University of Michigan

URL: http://hdl.handle.net/2027.42/110315

► To analyze the incentive structure of strategic multi-agent interactions, such scenarios are often cast as games, where players optimize their payoffs by selecting a strategy… (more)

▼ To analyze the incentive structure of strategic multi-agent interactions, such scenarios are often cast as games, where players optimize their payoffs by selecting a strategy in anticipation of the strategic decisions of other players. When our modeling needs are too complex to address analytically, empirical game models, game models in which observations of simulated play are used to estimate payoffs of agents, can be employed to facilitate game-theoretic analysis. This dissertation focuses on extending the capability of the empirical game-theoretic analysis (EGTA) framework for modeling and analyzing large games. My contributions are in three distinct areas: increasing the scale of game simulation through software infrastructure, improving performance of common analytic tasks by bringing them closer to the data, and reducing sampling requirements for statistically confident analysis through sequential sampling algorithms. With the advent of EGTAOnline, an experiment management system for distributed game simulation that I developed, EGTA practitioners no longer limit their studies to what can be conducted on a single computer. Over one billion payoff observations have been added to EGTAOnline's database to date, corresponding to hundreds of distinct experiments. To reduce the cost of analyzing this data, I explored conducting analysis in the database. I found that translating data to an in-memory object representation was a dominant cost for game-theoretic analysis software. By avoiding that cost, conducting analysis in the database improves performance. A further way to improve scalability is to ensure we only gather as much data as is necessary to support analysis. I developed algorithms that interweave sampling and evaluations of statistical confidence, improving on existing ad hoc sampling methods by providing a measure of statistical confidence for analysis and reducing the number of observations taken. In addition to these software and methodological contributions, I present two applications: a strategic analysis of selecting a wireless access point for your traffic, and an investigation of mapping an analytical pricing model to a large simulated stock market. Advisors/Committee Members: Wellman, Michael P. (committee member), MacKie-Mason, Jeffrey (committee member), Laird, John E. (committee member), Teneketzis, Demosthenis (committee member).

Subjects/Keywords: empirical game-theoretic analysis; scalability; distributed systems; sequential analysis; Economics; Computer Science; Statistics and Numeric Data; Business and Economics; Engineering; Science

10 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Cassell, B. (2014). Scaling Empirical Game-Theoretic Analysis. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/110315

Chicago Manual of Style (16^th Edition):

Cassell, Ben-Alexander. “Scaling Empirical Game-Theoretic Analysis.” 2014. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/110315.

MLA Handbook (7^th Edition):

Cassell, Ben-Alexander. “Scaling Empirical Game-Theoretic Analysis.” 2014. Web. 27 Jan 2021.

Vancouver:

Cassell B. Scaling Empirical Game-Theoretic Analysis. [Internet] [Doctoral dissertation]. University of Michigan; 2014. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/110315.

Council of Science Editors:

Cassell B. Scaling Empirical Game-Theoretic Analysis. [Doctoral Dissertation]. University of Michigan; 2014. Available from: http://hdl.handle.net/2027.42/110315

24. Wright, Mason. Stable Profiles in Simulation-Based Games via Reinforcement Learning and Statistics.

Degree: PhD, Computer Science & Engineering, 2019, University of Michigan

URL: http://hdl.handle.net/2027.42/149991

► In environments governed by the behavior of strategically interacting agents, game theory provides a way to predict outcomes in counterfactual scenarios, such as new market… (more)

▼ In environments governed by the behavior of strategically interacting agents, game theory provides a way to predict outcomes in counterfactual scenarios, such as new market mechanisms or cybersecurity systems. Simulation-based games allow analysts to reason about settings that are too complex to model analytically with sufficient fidelity. But prior techniques for studying agent behavior in simulation-based games lack theoretical guarantees about the strategic stability of these behaviors. In this dissertation, I propose a way to measure the likelihood an agent could find a beneficial strategy deviation from a proposed behavior, using a limited number of samples from a distribution over strategies, including a theoretically proven bound. This method employs a provably conservative confidence interval estimator, along with a multiple test correction, to provide its guarantee. I show that the method can reliably find provably stable strategy profiles in an auction game, and in a cybersecurity game from prior literature. I also present a method for evaluating the stability of strategy profiles learned over a restricted set of strategies, where a strategy profile is an assignment of a strategy to each agent in a game. This method uses reinforcement learning to challenge the learned behavior as a test of its soundness. This study finds that a widely-used trading agent model, the zero-intelligence trader, can be reasonably strategically stable in continuous double auction games, but only if the strategies have their parameters calibrated for the particular game instance. In addition, I present new applications of empirical game-theoretic analysis (EGTA) to a cybersecurity setting, involving defense against attacker intrusion into a computer system. This work uses iterated deep reinforcement learning to generate more strategically stable attacker and defender strategies, relative to those found in prior work. It also offers empirical insights into how iterated deep reinforcement learning approaches strategic equilibrium, over dozens of rounds. Advisors/Committee Members: Wellman, Michael P (committee member), Teneketzis, Demosthenis (committee member), Schoenebeck, Grant (committee member), Wiens, Jenna (committee member).

Subjects/Keywords: simulation-based games; reinforcement learning; game theory; Computer Science; Engineering

8 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Wright, M. (2019). Stable Profiles in Simulation-Based Games via Reinforcement Learning and Statistics. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/149991

Chicago Manual of Style (16^th Edition):

Wright, Mason. “Stable Profiles in Simulation-Based Games via Reinforcement Learning and Statistics.” 2019. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/149991.

MLA Handbook (7^th Edition):

Wright, Mason. “Stable Profiles in Simulation-Based Games via Reinforcement Learning and Statistics.” 2019. Web. 27 Jan 2021.

Vancouver:

Wright M. Stable Profiles in Simulation-Based Games via Reinforcement Learning and Statistics. [Internet] [Doctoral dissertation]. University of Michigan; 2019. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/149991.

Council of Science Editors:

Wright M. Stable Profiles in Simulation-Based Games via Reinforcement Learning and Statistics. [Doctoral Dissertation]. University of Michigan; 2019. Available from: http://hdl.handle.net/2027.42/149991

25. Du, Ye. Essays on the Computation of Economic Equilibria and Its Applications.

Degree: PhD, Computer Science & Engineering, 2009, University of Michigan

URL: http://hdl.handle.net/2027.42/64821

► The computation of economic equilibria is a central problem in algorithmic game theory. In this dissertation, we investigate the existence of economic equilibria in several… (more)

▼ The computation of economic equilibria is a central problem in algorithmic game theory. In this dissertation, we investigate the existence of economic equilibria in several markets and games, the complexity of computing economic equilibria, and its application to rankings. It is well known that a competitive economy always has an equilibrium under mild conditions. In this dissertation, we study the complexity of computing competitive equilibria. We show that given a competitive economy that fully respects all the conditions of Arrow-Debreu's existence theorem, it is PPAD-hard to compute an approximate competitive equilibrium. Furthermore, it is still PPAD-Complete to compute an approximate equilibrium for economies with additively separable piecewise linear concave utility functions. Degeneracy is an important concept in game theory. We study the complexity of deciding degeneracy in games. We show that it is NP-Complete to decide whether a bimatrix game is degenerate. With the advent of the Internet, an agent can easily have access to multiple accounts. In this dissertation we study the path auction game, which is a model for QoS routing, supply chain management, and so on, with multiple edge ownership. We show that the condition of multiple edge ownership eliminates the possibility of reasonable solution concepts, such as a strategyproof or false-name-proof mechanism or Pareto efficient Nash equilibria. The stationary distribution (an equilibrium point) of a Markov chain is widely used for ranking purposes. One of the most important applications is PageRank, part of the ranking algorithm of Google. By making use of perturbation theories of Markov chains, we show the optimal manipulation strategies of a Web spammer against PageRank under a few natural constraints. Finally, we make a connection between the ranking vector of PageRank or the Invariant method and the equilibrium of a Cobb-Douglas market. Furthermore, we propose the CES ranking method based on the Constant Elasticity of Substitution (CES) utility functions. Advisors/Committee Members: Shi, Yaoyun (committee member), Saigal, Romesh (committee member), Sami, Rahul (committee member), Wellman, Michael P. (committee member).

Subjects/Keywords: Competitive Equilibrium; PPAD; Path Auction; PageRank; Computer Science; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Du, Y. (2009). Essays on the Computation of Economic Equilibria and Its Applications. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/64821

Chicago Manual of Style (16^th Edition):

Du, Ye. “Essays on the Computation of Economic Equilibria and Its Applications.” 2009. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/64821.

MLA Handbook (7^th Edition):

Du, Ye. “Essays on the Computation of Economic Equilibria and Its Applications.” 2009. Web. 27 Jan 2021.

Vancouver:

Du Y. Essays on the Computation of Economic Equilibria and Its Applications. [Internet] [Doctoral dissertation]. University of Michigan; 2009. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/64821.

Council of Science Editors:

Du Y. Essays on the Computation of Economic Equilibria and Its Applications. [Doctoral Dissertation]. University of Michigan; 2009. Available from: http://hdl.handle.net/2027.42/64821

26. Witwicki, Stefan J. Abstracting Influences for Efficient Multiagent Coordination Under Uncertainty.

Degree: PhD, Computer Science & Engineering, 2011, University of Michigan

URL: http://hdl.handle.net/2027.42/84614

► When planning optimal decisions for teams of agents acting in uncertain domains, conventional methods explicitly coordinate all joint policy decisions and, in doing so, are… (more)

▼ When planning optimal decisions for teams of agents acting in uncertain domains, conventional methods explicitly coordinate all joint policy decisions and, in doing so, are inherently susceptible to the curse of dimensionality, as state, action, and observation spaces grow exponentially with the number of agents. With the goal of extending the scalability of optimal team coordination, the research presented in this dissertation examines how agents can reduce the amount of information they need to coordinate. Intuitively, to the extent that agents are weakly coupled, they can avoid the complexity of coordinating all decisions; they need instead only coordinate abstractions of their policies that convey their essential influences on each other. In formalizing this intuition, I consider several complementary aspects of weakly-coupled problem structure, including agent scope size, corresponding to the number of an agent's peers whose decisions influence the agent's decisions, and degree of influence, corresponding to the proportion of unique influences that peers can feasibly exert. To exploit this structure, I introduce a (transition-dependent decentralized POMDP) model that efficiently decomposes into local decision models with shared state features. This context yields a novel characterization of influences as transition probabilities (compactly encoded using a dynamic Bayesian network). Not only is this influence representation provably sufficient for optimal coordination, but it also allows me to frame the subproblems of (1) proposing influences, (2) evaluating influences, and (3) computing optimal policies around influences as mixed-integer linear programs. The primary advantage of working in the influence space is that there are potentially significantly fewer feasible influences than there are policies. Blending prior work on decoupled joint policy search and constraint optimization, I develop influence-space search algorithms that, for problems with a low degree of influence, compute optimal solutions orders of magnitude faster than policy-space search. When agents' influences are constrained, influence-space search also outperforms other state-of-the-art optimal solution algorithms. Moreover, by exploiting both degree of influence and agent scope size, I demonstrate scalability, substantially beyond the reach of prior optimal methods, to teams of 50 weakly-coupled transition-dependent agents. Advisors/Committee Members: Durfee, Edmund H. (committee member), Baveja, Satinder Singh (committee member), Cohn, Amy Ellen (committee member), Wellman, Michael P. (committee member).

Subjects/Keywords: Multiagent Coordination; Optimal Planning; Sequential Decision Making; Agent Interaction Structure; Decentralized Partially Observable Markov Decision Process; Influence Based Policy Abstraction; Computer Science; Engineering

10 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Witwicki, S. J. (2011). Abstracting Influences for Efficient Multiagent Coordination Under Uncertainty. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/84614

Chicago Manual of Style (16^th Edition):

Witwicki, Stefan J. “Abstracting Influences for Efficient Multiagent Coordination Under Uncertainty.” 2011. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/84614.

MLA Handbook (7^th Edition):

Witwicki, Stefan J. “Abstracting Influences for Efficient Multiagent Coordination Under Uncertainty.” 2011. Web. 27 Jan 2021.

Vancouver:

Witwicki SJ. Abstracting Influences for Efficient Multiagent Coordination Under Uncertainty. [Internet] [Doctoral dissertation]. University of Michigan; 2011. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/84614.

Council of Science Editors:

Witwicki SJ. Abstracting Influences for Efficient Multiagent Coordination Under Uncertainty. [Doctoral Dissertation]. University of Michigan; 2011. Available from: http://hdl.handle.net/2027.42/84614

27. Martin, Travis Bennett. Theoretical Tools for Network Analysis: Game Theory, Graph Centrality, and Statistical Inference.

Degree: PhD, Computer Science and Engineering, 2016, University of Michigan

URL: http://hdl.handle.net/2027.42/133463

► A computer-driven data explosion has made the difficulty of interpreting large data sets of interconnected entities ever more salient. My work focuses on theoretical tools… (more)

▼ A computer-driven data explosion has made the difficulty of interpreting large data sets of interconnected entities ever more salient. My work focuses on theoretical tools for summarizing, analyzing, and understanding network data sets, or data sets of things and their pairwise connections. I address four network science issues, improving our ability to analyze networks from a variety of domains. I first show that the sophistication of game-theoretic agent decision making can crucially effect network cascades: differing decision making assumptions can lead to dramatically different cascade outcomes. This highlights the importance of diligence when making assumptions about agent behavior on networks and in general. I next analytically demonstrate a significant irregularity in the popular eigenvector centrality, and propose a new spectral centrality measure, nonbacktracking centrality, showing that it avoids this irregularity. This tool contributes a more robust way of ranking nodes, as well as an additional mathematical understanding of the effects of network localization. I next give a new model for uncertain networks, networks in which one has no access to true network data but instead observes only probabilistic information about edge existence. I give a fast maximum-likelihood algorithm for recovering edges and communities in this model, and show that it outperforms a typical approach of thresholding to an unweighted network. This model gives a better tool for understanding and analyzing real-world uncertain networks such as those arising in the experimental sciences. Lastly, I give a new lens for understanding scientific literature, specifically as a hybrid coauthorship and citation network. I use this for exploratory analysis of the Physical Review journals over a hundred-year period, and I make new observations about the interplay between these two networks and how this relationship has changed over time. Advisors/Committee Members: Newman, Mark E (committee member), Wellman, Michael P. (committee member), Nadakuditi, Rajesh Rao (committee member), Pettie, Seth (committee member), Schoenebeck, Grant (committee member).

Subjects/Keywords: Network science; Computer Science; Engineering

…network of all college students, students at University of Michigan (UM) display…

13 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Martin, T. B. (2016). Theoretical Tools for Network Analysis: Game Theory, Graph Centrality, and Statistical Inference. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/133463

Chicago Manual of Style (16^th Edition):

Martin, Travis Bennett. “Theoretical Tools for Network Analysis: Game Theory, Graph Centrality, and Statistical Inference.” 2016. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/133463.

MLA Handbook (7^th Edition):

Martin, Travis Bennett. “Theoretical Tools for Network Analysis: Game Theory, Graph Centrality, and Statistical Inference.” 2016. Web. 27 Jan 2021.

Vancouver:

Martin TB. Theoretical Tools for Network Analysis: Game Theory, Graph Centrality, and Statistical Inference. [Internet] [Doctoral dissertation]. University of Michigan; 2016. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/133463.

Council of Science Editors:

Martin TB. Theoretical Tools for Network Analysis: Game Theory, Graph Centrality, and Statistical Inference. [Doctoral Dissertation]. University of Michigan; 2016. Available from: http://hdl.handle.net/2027.42/133463

28. Dimitrov, Stanko B. Information Procurement and Delivery: Robustness in Prediction Markets and Network Routing.

Degree: PhD, Industrial & Operations Engineering, 2010, University of Michigan

URL: http://hdl.handle.net/2027.42/75962

► In this dissertation we address current problems in information procurement and delivery. Uncertainty commonly reduces the efficacy of information procurement systems, such as prediction markets,… (more)

▼ In this dissertation we address current problems in information procurement and delivery. Uncertainty commonly reduces the efficacy of information procurement systems, such as prediction markets, and information delivery systems, such as Internet backbone networks. We address the problems of uncertainty by designing robust algorithms and protocols that function well under uncertainty. Telecommunication backbone networks are used for delivering information across the Internet. Current backbone networks mostly employ protocols that include sender-receiver based congestion control. However, as protocols that do not have congestion control available become more prevalent, the network routers themselves must perform congestion control. In order to maximize network throughput, routing policies for backbone networks that take into account router based congestion control must be devised. We propose a mathematical model that can be used to design improved routing policies, while also taking into account existing flow management methods. Our model incorporates current active congestion control methods, and takes into account demand uncertainty when creating routing policies. The resulting routing policies tended to be at least 20% better than those currently used in a real world network in our experiments. Prediction markets are information aggregation tools in which participants trade on the outcome of a future event. One commonly used form of prediction market, the market scoring rule market, accurately aggregate the beliefs of traders assuming the traders are myopic, meaning they do not consider future payoffs, and are risk neutral. In currently deployed prediction markets neither of these assumptions typically holds. Therefore, in order to analyze the effectiveness of such markets, we look at the impact of non-myopic risk neutral traders, as well as risk averse traders on prediction markets. We identify a setting where non-myopic risk neutral traders may bluff, and propose a modified prediction market to disincentivize such behavior. Current prediction markets do not accurately aggregate all risk averse traders' beliefs. Therefore, we propose a new prediction market that does. The resulting market exponentially reduces the reward given to traders as the number of traders increases; we show that this exponential reduction is necessary for any prediction market that aggregates the beliefs of risk averse traders. Advisors/Committee Members: Epelman, Marina A. (committee member), Sami, Rahul (committee member), Keppo, Jussi Samuli (committee member), Sharma, Dushyant (committee member), Wellman, Michael P. (committee member).

Subjects/Keywords: Network Routing; Prediction Markets; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Dimitrov, S. B. (2010). Information Procurement and Delivery: Robustness in Prediction Markets and Network Routing. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/75962

Chicago Manual of Style (16^th Edition):

Dimitrov, Stanko B. “Information Procurement and Delivery: Robustness in Prediction Markets and Network Routing.” 2010. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/75962.

MLA Handbook (7^th Edition):

Dimitrov, Stanko B. “Information Procurement and Delivery: Robustness in Prediction Markets and Network Routing.” 2010. Web. 27 Jan 2021.

Vancouver:

Dimitrov SB. Information Procurement and Delivery: Robustness in Prediction Markets and Network Routing. [Internet] [Doctoral dissertation]. University of Michigan; 2010. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/75962.

Council of Science Editors:

Dimitrov SB. Information Procurement and Delivery: Robustness in Prediction Markets and Network Routing. [Doctoral Dissertation]. University of Michigan; 2010. Available from: http://hdl.handle.net/2027.42/75962

29. Naghizadeh Ardabili, Parinaz. On the Provision of Public Goods on Networks: Incentives, Exit Equilibrium, and Applications to Cyber .

Degree: PhD, Electrical Engineering: Systems, 2016, University of Michigan

URL: http://hdl.handle.net/2027.42/133328

► Attempts to improve the state of cyber-security have been on the rise over the past years. The importance of incentivizing better security decisions by users… (more)

▼ Attempts to improve the state of cyber-security have been on the rise over the past years. The importance of incentivizing better security decisions by users in the current landscape is two-fold: it not only helps users protect themselves against attacks, but also provides positive externalities to others interacting with them, as a protected user is less likely to become compromised and be used to propagate attacks against other entities. Therefore, security can be viewed as a public good. This thesis takes a game-theoretic approach to understanding the theoretical underpinnings of users' incentives in the provision of public goods, and in particular, cyber-security. We analyze the strategic interactions of users in the provision of security as a non-excludable public good. We propose the notion of exit equilibrium to describe users' outside options from mechanisms for incentivizing the adoption of better security decisions, and use it to highlight the crucial effect of outside options on the design of incentive mechanisms for improving the state of cyber-security. We further focus on the general problem of public good provision games on networks. We identify necessary and sufficient conditions on the structure of the network for the existence and uniqueness of the Nash equilibrium in these games. We show that previous results in the literature can be recovered as special cases of our result. We provide a graph-theoretical interpretation of users' efforts at the Nash equilibria, Pareto efficient outcomes, and semi-cooperative equilibria of these games, by linking users' effort decisions to their centralities in the interaction network. Using this characterization, we separate the effects of users' dependencies and influences (outgoing and incoming edges, respectively) on their effort levels, and uncover an alternating effect over walks of different length in the network. We also propose the design of inter-temporal incentives in a particular type of security games, namely, security information sharing agreement. We show that either public or private assessments can be used in designing incentives for participants to disclose their information in these agreements. Finally, we present a method for crowdsourcing reputation that can be useful in attaining assessments of users' efforts in security games. Advisors/Committee Members: Liu, Mingyan (committee member), Wellman, Michael P (committee member), Teneketzis, Demosthenis (committee member), Subramanian, Vijay Gautam (committee member), Loeb, Martin P (committee member).

Subjects/Keywords: Cyber Security; Network Games; Public Goods; Game Theory; Economics; Computer Science; Electrical Engineering; Business and Economics; Engineering

7 more images…

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Naghizadeh Ardabili, P. (2016). On the Provision of Public Goods on Networks: Incentives, Exit Equilibrium, and Applications to Cyber . (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/133328

Chicago Manual of Style (16^th Edition):

Naghizadeh Ardabili, Parinaz. “On the Provision of Public Goods on Networks: Incentives, Exit Equilibrium, and Applications to Cyber .” 2016. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/133328.

MLA Handbook (7^th Edition):

Naghizadeh Ardabili, Parinaz. “On the Provision of Public Goods on Networks: Incentives, Exit Equilibrium, and Applications to Cyber .” 2016. Web. 27 Jan 2021.

Vancouver:

Naghizadeh Ardabili P. On the Provision of Public Goods on Networks: Incentives, Exit Equilibrium, and Applications to Cyber . [Internet] [Doctoral dissertation]. University of Michigan; 2016. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/133328.

Council of Science Editors:

Naghizadeh Ardabili P. On the Provision of Public Goods on Networks: Incentives, Exit Equilibrium, and Applications to Cyber . [Doctoral Dissertation]. University of Michigan; 2016. Available from: http://hdl.handle.net/2027.42/133328

30. Brinkman, Erik. Understanding Financial Market Behavior through Empirical Game-Theoretic Analysis.

Degree: PhD, Computer Science & Engineering, 2018, University of Michigan

URL: http://hdl.handle.net/2027.42/144031

► Financial market activity is increasingly controlled by algorithms, interacting through electronic markets. Unprecedented information response times, autonomous operation, use of machine learning and other adaptive… (more)

▼ Financial market activity is increasingly controlled by algorithms, interacting through electronic markets. Unprecedented information response times, autonomous operation, use of machine learning and other adaptive techniques, and ability to proliferate novel strategies at scale are all reasons to question whether algorithmic trading may produce dynamic behavior qualitatively different from what arises in trading under direct human control. Given the high level of competition between trading firms and the significant financial incentives to trading, it is desirable to understand the effect incentives have on the behavior of agents in financial markets. One natural way to analyze this effect is through the economic concept of a Nash equilibrium, a behavior profile of every agent such that no individual stands to gain by doing something different. Some of the incentives traders face arise from the complexities of modern market structure. Recent studies have turned to agent-based modeling as a way to capture behavioral response to this structure. Agent-based modeling is a simulation paradigm that allows studying the interaction of agents in a simulated environment, and it has been used to model various aspects of financial market structure. This thesis builds on recent agent-based models of financial markets by imposing agent rationality and studying the models in equilibrium. I use empirical game-theoretic analysis, a methodology for computing approximately rational behavior in agent-based models, to investigate three important aspects of market structure. First, I evaluate the impact of strategic bid shading on agent welfare. Bid shading is when agents demand better prices, lower if they are buying or higher if they are selling, and is typically associated with lower social welfare. My results indicate that in many market environments, strategic bid shading actually improves social welfare, even with some of the complexities of financial markets. Next, I investigate the optimal clearing interval for a proposed market mechanism, the frequent call market. There is significant evidence to support the idea that traders will benefit from trading in a frequent call market over standard continuous double auction markets. My results confirm this statement for a wide variety of market settings, but I also find a few circumstances, particularly when large informational advantages exist, or when markets are thin, that call markets consistently hurt welfare, independent of frequency. I conclude with an investigation on the effect of trend following on market stability. Here I find that the presence of trend followers alters a market’s response to shock. In the absence of trend followers, shocks are small but have a long recovery. When trend followers are present, they alter background trader behavior resulting in more severe shocks that recover much more quickly. I also develop a novel method to efficiently evaluate the effect of shock anticipation on equilibrium. While anticipation of shocks does make markets more stable,… Advisors/Committee Members: Wellman, Michael P (committee member), Rajan, Uday (committee member), Abernethy, Jacob (committee member), Page, Scott E (committee member), Sinha, Arunesh (committee member).

Subjects/Keywords: empirical game-theoretic analysis; financial markets; Computer Science; Engineering

Record Details Similar Records Cite « Share

❌

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6^th Edition):

Brinkman, E. (2018). Understanding Financial Market Behavior through Empirical Game-Theoretic Analysis. (Doctoral Dissertation). University of Michigan. Retrieved from http://hdl.handle.net/2027.42/144031

Chicago Manual of Style (16^th Edition):

Brinkman, Erik. “Understanding Financial Market Behavior through Empirical Game-Theoretic Analysis.” 2018. Doctoral Dissertation, University of Michigan. Accessed January 27, 2021. http://hdl.handle.net/2027.42/144031.

MLA Handbook (7^th Edition):

Brinkman, Erik. “Understanding Financial Market Behavior through Empirical Game-Theoretic Analysis.” 2018. Web. 27 Jan 2021.

Vancouver:

Brinkman E. Understanding Financial Market Behavior through Empirical Game-Theoretic Analysis. [Internet] [Doctoral dissertation]. University of Michigan; 2018. [cited 2021 Jan 27]. Available from: http://hdl.handle.net/2027.42/144031.

Council of Science Editors:

Brinkman E. Understanding Financial Market Behavior through Empirical Game-Theoretic Analysis. [Doctoral Dissertation]. University of Michigan; 2018. Available from: http://hdl.handle.net/2027.42/144031

		in
And / Or
		in
And / Or
		in
And / Or
		in

Open Access Theses and Dissertations