subject:(Reinforcement Learning)

Oregon State University

1. Proper, Scott. Scaling multiagent reinforcement learning.

Degree: PhD, Computer Science, 2009, Oregon State University

URL: http://hdl.handle.net/1957/13662

► Reinforcement learning in real-world domains suffers from three curses of dimensionality: explosions in state and action spaces, and high stochasticity or "outcome space" explosion. Multiagent… (more)

▼ Reinforcement learning in real-world domains suffers from three curses of dimensionality: explosions in state and action spaces, and high stochasticity or "outcome space" explosion. Multiagent domains are particularly susceptible to these problems. This thesis describes ways to mitigate these curses in several different multiagent domains, including real-time delivery of products using multiple vehicles with stochastic demands, a multiagent predator-prey domain, and a domain based on a real-time strategy game. To mitigate the problem of state-space explosion, this thesis present several approaches that mitigate each of these curses. "Tabular linear functions" (TLFs) are introduced that generalize tile-coding and linear value functions and allow learning of complex nonlinear functions in high-dimensional state-spaces. It is also shown how to adapt TLFs to relational domains, creating a "lifted" version called relational templates. To mitigate the problem of action-space explosion, the replacement of complete joint action space search with a form of hill climbing is described. To mitigate the problem of outcome space explosion, a more efficient calculation of the expected value of the next state is shown, and two real-time dynamic programming algorithms based on afterstates, ASH-learning and ATR-learning, are introduced. Lastly, two approaches that scale by treating a multiagent domain as being formed of several coordinating agents are presented. "Multiagent H-learning" and "Multiagent ASH-learning" are described, where coordination is achieved through a method called "serial coordination". This technique has the benefit of addressing each of the three curses of dimensionality simultaneously by reducing the space of states and actions each local agent must consider. The second approach to multiagent coordination presented is "assignment-based decomposition", which divides the action selection step into an assignment phase and a primitive action selection step. Like the multiagent approach, assignment-based decomposition addresses all three curses of dimensionality simultaneously by reducing the space of states and actions each group of agents must consider. This method is capable of much more sophisticated coordination. Experimental results are presented which show successful application of all methods described. These results demonstrate that the scaling techniques described in this thesis can greatly mitigate the three curses of dimensionality and allow solutions for multiagent domains to scale to large numbers of agents, and complex state and outcome spaces. Advisors/Committee Members: Tadepalli, Prasad (advisor), Dietterich, Thomas (committee member).

Subjects/Keywords: Reinforcement learning; Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Proper, S. (2009). Scaling multiagent reinforcement learning. (Doctoral Dissertation). Oregon State University. Retrieved from http://hdl.handle.net/1957/13662

Chicago Manual of Style (16^th Edition):

Proper, Scott. “Scaling multiagent reinforcement learning.” 2009. Doctoral Dissertation, Oregon State University. Accessed March 04, 2021. http://hdl.handle.net/1957/13662.

MLA Handbook (7^th Edition):

Proper, Scott. “Scaling multiagent reinforcement learning.” 2009. Web. 04 Mar 2021.

Vancouver:

Proper S. Scaling multiagent reinforcement learning. [Internet] [Doctoral dissertation]. Oregon State University; 2009. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/1957/13662.

Council of Science Editors:

Proper S. Scaling multiagent reinforcement learning. [Doctoral Dissertation]. Oregon State University; 2009. Available from: http://hdl.handle.net/1957/13662

Oregon State University

2. Mehta, Neville. Hierarchical structure discovery and transfer in sequential decision problems.

Degree: PhD, Computer Science, 2011, Oregon State University

URL: http://hdl.handle.net/1957/25199

► Acting intelligently to efficiently solve sequential decision problems requires the ability to extract hierarchical structure from the underlying domain dynamics, exploit it for optimal or… (more)

▼ Acting intelligently to efficiently solve sequential decision problems requires the ability to extract hierarchical structure from the underlying domain dynamics, exploit it for optimal or near-optimal decision-making, and transfer it to related problems instead of solving every problem in isolation. This dissertation makes three contributions toward this goal. The first contribution is the introduction of two frameworks for the transfer of hierarchical structure in sequential decision problems. The MASH framework facilitates transfer among multiple agents coordinating within a domain. The VRHRL framework allows an agent to transfer its knowledge across a family of domains that share the same transition dynamics but have differing reward dynamics. Both MASH and VRHRL are validated empirically in large domains and the results demonstrate significant speedup in the solutions due to transfer. The second contribution is a new approach to the discovery of hierarchical structure in sequential decision problems. HI-MAT leverages action models to analyze the relevant dependencies in a hierarchically-generated trajectory and it discovers hierarchical structure that transfers to all problems whose actions share the same relevant dependencies as the single source problem. HierGen advances HI-MAT by learning simple action models, leveraging these models to analyze non-hierarchically-generated trajectories from multiple source problems in a robust causal fashion, and discovering hierarchical structure that transfers to all problems whose actions share the same causal dependencies as those in the source problems. Empirical evaluations in multiple domains demonstrate that the discovered hierarchical structures are comparable to manually-designed structures in quality and performance. Action models are essential to hierarchical structure discovery and other aspects of intelligent behavior. The third contribution of this dissertation is the introduction of two general frameworks for learning action models in sequential decision problems. In the MBP framework, learning is user-driven; in the PLEX framework, the learner generates its own problems. The frameworks are formally analyzed and reduced to concept learning with one-sided error. A general action-modeling language is shown to be efficiently learnable in both frameworks. Advisors/Committee Members: Tadepalli, Prasad (advisor), Dietterich, Thomas (committee member).

Subjects/Keywords: hierarchical reinforcement learning; Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Mehta, N. (2011). Hierarchical structure discovery and transfer in sequential decision problems. (Doctoral Dissertation). Oregon State University. Retrieved from http://hdl.handle.net/1957/25199

Chicago Manual of Style (16^th Edition):

Mehta, Neville. “Hierarchical structure discovery and transfer in sequential decision problems.” 2011. Doctoral Dissertation, Oregon State University. Accessed March 04, 2021. http://hdl.handle.net/1957/25199.

MLA Handbook (7^th Edition):

Mehta, Neville. “Hierarchical structure discovery and transfer in sequential decision problems.” 2011. Web. 04 Mar 2021.

Vancouver:

Mehta N. Hierarchical structure discovery and transfer in sequential decision problems. [Internet] [Doctoral dissertation]. Oregon State University; 2011. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/1957/25199.

Council of Science Editors:

Mehta N. Hierarchical structure discovery and transfer in sequential decision problems. [Doctoral Dissertation]. Oregon State University; 2011. Available from: http://hdl.handle.net/1957/25199

Oregon State University

3. Lauer, Christopher Joseph. Determining optimal timber harvest and fuel treatment on a fire-threatened landscape using approximate dynamic programming.

Degree: PhD, 2017, Oregon State University

URL: http://hdl.handle.net/1957/61678

► Forest management in the face of fire risk is a challenging problem because fire spreads across a landscape and because its occurrence is unpredictable. Additionally,… (more)

▼ Forest management in the face of fire risk is a challenging problem because fire spreads across a landscape and because its occurrence is unpredictable. Additionally, management must be adjusted over time as unpredictable fire events are realized. Land managers have some control over the vegetation conditions that facilitate fire spread, and they can engage in activity that reduces fire arrival probability and decreases the likelihood of damage. This management activity will have consequences for fire risk in multiple locations on a landscape over a long time-horizon. The first manuscript in this dissertation demonstrates a method for incorporating spatial information and interactions into a dynamic model for management decisions. Approximate dynamic programming is applied to determine the optimal timing and location of fuel treatments and timber harvests for a fire-threatened landscape. This method allows a value function and an optimal policy to be found for a problem with high dimension state and action spaces. Larger net present values can be achieved using policies that explicitly consider evolving spatial interactions created by fire spread, compared to policies that are optimal in a non-spatial or non-dynamic setting. The second manuscript explores the interaction of multiple landowners on a single landscape. Fuel conditions that allow fire to cross property boundaries creates externalities, and must be accounted for when making management decisions. In the second manuscript, the problem of determining optimal management is formulated as a dynamic game where each agent affects the fire risk for other agents on the landscape. Value functions and optimal policies that account for fire-spread externalities are determined for each agent. This method is applied to analyze the effect of ownership fragmentation on landowner welfare and ecological outcomes. The third manuscript explores policies designed to make landowners accountable for externalities generated by their management. Damage caused by spreading fires has prompted the creation of laws such as liability and negligence standards for landowners. Large damaging fires that resulted in expensive litigation have prompted changes to wildfire liability laws in several states. These regulations change the incentives that managers face. In this manuscript, the multi-agent framework developed in the second part of this dissertation is expanded and applied to analyze the effect of different types of liability regulations. Advisors/Committee Members: Montgomery, Claire A. (advisor), Dietterich, Thomas G. (committee member).

Subjects/Keywords: reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Lauer, C. J. (2017). Determining optimal timber harvest and fuel treatment on a fire-threatened landscape using approximate dynamic programming. (Doctoral Dissertation). Oregon State University. Retrieved from http://hdl.handle.net/1957/61678

Chicago Manual of Style (16^th Edition):

Lauer, Christopher Joseph. “Determining optimal timber harvest and fuel treatment on a fire-threatened landscape using approximate dynamic programming.” 2017. Doctoral Dissertation, Oregon State University. Accessed March 04, 2021. http://hdl.handle.net/1957/61678.

MLA Handbook (7^th Edition):

Lauer, Christopher Joseph. “Determining optimal timber harvest and fuel treatment on a fire-threatened landscape using approximate dynamic programming.” 2017. Web. 04 Mar 2021.

Vancouver:

Lauer CJ. Determining optimal timber harvest and fuel treatment on a fire-threatened landscape using approximate dynamic programming. [Internet] [Doctoral dissertation]. Oregon State University; 2017. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/1957/61678.

Council of Science Editors:

Lauer CJ. Determining optimal timber harvest and fuel treatment on a fire-threatened landscape using approximate dynamic programming. [Doctoral Dissertation]. Oregon State University; 2017. Available from: http://hdl.handle.net/1957/61678

4. Frank, Mikhail Alexander. Learning to reach and reaching to learn: a unified approach to path planning and reactive control through reinforcement learning.

Degree: 2014, Università della Svizzera italiana

URL: http://doc.rero.ch/record/234387

► The next generation of intelligent robots will need to be able to plan reaches. Not just ballistic point to point reaches, but reaches around things… (more)

▼ The next generation of intelligent robots will need to be able to plan reaches. Not just ballistic point to point reaches, but reaches around things such as the edge of a table, a nearby human, or any other known object in the robot’s workspace. Planning reaches may seem easy to us humans, because we do it so intuitively, but it has proven to be a challenging problem, which continues to limit the versatility of what robots can do today. In this document, I propose a novel intrinsically motivated RL system that draws on both Path/Motion Planning and Reactive Control. Through Reinforcement Learning, it tightly integrates these two previously disparate approaches to robotics. The RL system is evaluated on a task, which is as yet unsolved by roboticists in practice. That is to put the palm of the iCub humanoid robot on arbitrary target objects in its workspace, start- ing from arbitrary initial configurations. Such motions can be generated by planning, or searching the configuration space, but this typically results in some kind of trajectory, which must then be tracked by a separate controller, and such an approach offers a brit- tle runtime solution because it is inflexible. Purely reactive systems are robust to many problems that render a planned trajectory infeasible, but lacking the capacity to search, they tend to get stuck behind constraints, and therefore do not replace motion planners. The planner/controller proposed here is novel in that it deliberately plans reaches without the need to track trajectories. Instead, reaches are composed of sequences of reactive motion primitives, implemented by my Modular Behavioral Environment (MoBeE), which provides (fictitious) force control with reactive collision avoidance by way of a realtime kinematic/geometric model of the robot and its workspace. Thus, to the best of my knowledge, mine is the first reach planning approach to simultaneously offer the best of both the Path/Motion Planning and Reactive Control approaches. By controlling the real, physical robot directly, and feeling the influence of the con- straints imposed by MoBeE, the proposed system learns a stochastic model of the iCub’s configuration space. Then, the model is exploited as a multiple query path planner to find sensible pre-reach poses, from which to initiate reaching actions. Experiments show that the system can autonomously find practical reaches to target objects in workspace and offers excellent robustness to changes in the workspace configuration as well as noise in the robot’s sensory-motor apparatus. Advisors/Committee Members: Schmidhuber, Jürgen (Dir.), Alexander (Codir.).

Subjects/Keywords: Reinforcement learning

11 more images…

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Frank, M. A. (2014). Learning to reach and reaching to learn: a unified approach to path planning and reactive control through reinforcement learning. (Thesis). Università della Svizzera italiana. Retrieved from http://doc.rero.ch/record/234387

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Frank, Mikhail Alexander. “Learning to reach and reaching to learn: a unified approach to path planning and reactive control through reinforcement learning.” 2014. Thesis, Università della Svizzera italiana. Accessed March 04, 2021. http://doc.rero.ch/record/234387.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Frank, Mikhail Alexander. “Learning to reach and reaching to learn: a unified approach to path planning and reactive control through reinforcement learning.” 2014. Web. 04 Mar 2021.

Vancouver:

Frank MA. Learning to reach and reaching to learn: a unified approach to path planning and reactive control through reinforcement learning. [Internet] [Thesis]. Università della Svizzera italiana; 2014. [cited 2021 Mar 04]. Available from: http://doc.rero.ch/record/234387.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Frank MA. Learning to reach and reaching to learn: a unified approach to path planning and reactive control through reinforcement learning. [Thesis]. Università della Svizzera italiana; 2014. Available from: http://doc.rero.ch/record/234387

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Delft University of Technology

5. Tian, Yuan (author). Model Free Reinforcement Learning with Stability Guarantee.

Degree: 2019, Delft University of Technology

URL: http://resolver.tudelft.nl/uuid:dde4e58f-e109-4e7f-8ecb-ed1734294e5c

►

Model-free reinforcement learning has proved to be successful in many tasks such as robotic manipulator, video games, and even stock trading. However, as the dynamics… (more)

▼

Model-free reinforcement learning has proved to be successful in many tasks such as robotic manipulator, video games, and even stock trading. However, as the dynamics of the environment is unmodelled, it is fundamentally difficult to ensure the learned policy to be absolutely reliable and its performance is guaranteed. In this thesis, we borrow the concept of stability and Lyapunov analysis in control theory to design a policy with stability guarantee and assure the guaranteed behaviors of the agent. A novel sample-based approach is proposed for analyzing the stability of a learning control system, and on the basis of the theoretical result, we establish a practical model-free learning framework with provable stability, safety and performance guarantees. % Specifically, a novel locally constrained method is proposed to solve the safety constrained problems with lower conservatism. In our solution, a Lyapunov function is searched automatically to guarantee the closed-loop system stability, which also guides the simultaneous learning (covering both the policy and value-based learning methods). Our approach is evaluated on a series of discrete and continuous control benchmarks and largely outperforms the state-of-the-art results concerning unconstrained and constrained problems. It is also shown that the algorithm has the ability of recovery to equilibrium under perturbation using the policy with stability guarantee. (Anonymous code is available to reproduce the experimental esults\footnote{\url{https://github.com/RLControlTheoreticGuarantee/Guarantee_Learning_Control}}.) Since sometimes the constraint is hard to define, we introduce a novel method to learn a constraint by representing the bad cases or situations as a distribution, and the constraint is the Wasserstein distance between the distribution.

Mechanical Engineering | Vehicle Engineering

Advisors/Committee Members: Pan, Wei (mentor), Zhou, Hongpeng (graduation committee), Delft University of Technology (degree granting institution).

Subjects/Keywords: Reinforcement Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Tian, Y. (. (2019). Model Free Reinforcement Learning with Stability Guarantee. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:dde4e58f-e109-4e7f-8ecb-ed1734294e5c

Chicago Manual of Style (16^th Edition):

Tian, Yuan (author). “Model Free Reinforcement Learning with Stability Guarantee.” 2019. Masters Thesis, Delft University of Technology. Accessed March 04, 2021. http://resolver.tudelft.nl/uuid:dde4e58f-e109-4e7f-8ecb-ed1734294e5c.

MLA Handbook (7^th Edition):

Tian, Yuan (author). “Model Free Reinforcement Learning with Stability Guarantee.” 2019. Web. 04 Mar 2021.

Vancouver:

Tian Y(. Model Free Reinforcement Learning with Stability Guarantee. [Internet] [Masters thesis]. Delft University of Technology; 2019. [cited 2021 Mar 04]. Available from: http://resolver.tudelft.nl/uuid:dde4e58f-e109-4e7f-8ecb-ed1734294e5c.

Council of Science Editors:

Tian Y(. Model Free Reinforcement Learning with Stability Guarantee. [Masters Thesis]. Delft University of Technology; 2019. Available from: http://resolver.tudelft.nl/uuid:dde4e58f-e109-4e7f-8ecb-ed1734294e5c

Delft University of Technology

6. Van Rooijen, J.C. (author). Learning Parameter Selection in Continuous Reinforcement Learning: Attempting to Reduce Tuning Effords.

Degree: 2012, Delft University of Technology

URL: http://resolver.tudelft.nl/uuid:94b81bc2-aff6-457f-9b54-be5e005def38

►

The reinforcement learning (RL) framework enables to construct controllers that try to find find an optimal control strategy in an unknown environment by trial-and-error. After… (more)

▼

The reinforcement learning (RL) framework enables to construct controllers that try to find find an optimal control strategy in an unknown environment by trial-and-error. After selecting a control action, the controller receives a numerical reward. The reward signal is based on the current state of the environment and the applied control action. The controller aims to maximize the cumulative reward, known as the return. In this thesis actor-critic and criticonly RL algorithms are considered. Actor-critic algorithms consist of an element that selects the actions (the actor) and an element that learns the expectation of the return (the critic). This expectation is captured in a value function. The critic is used to improve the control policy of the actor. Critic-only algorithms select the action by direct optimization over a value function. Before a RL algorithm can be applied to a control problem, a number of learning parameters need to be set. The optimal values of some of these parameters are highly problem dependent. It is not straightforward how these parameters should be chosen and often these are often determined by trying a large set of parameters. The main focus of thesis is to devise an action selection method that is able to select continuous actions without problem dependent parameters. Two approaches are taken: first, it is investigated if Levenberg Marquardt (LM), a popular optimization method, can be used to determine the actor update step. Second, an action selection method is treated that lacks an explicit actor, called Value-Gradient Based Policy (VGBP). The LM algorithm uses the gradient and the Hessian to compute the update step. Therefore the policy gradient and Hessian need to be found. A novel actor-critic method has been devised, called Vanilla Actor-Critic (VAC), that efficiently learns the policy gradient. On the inverted pendulum swing-up task this algorithm outperformed Natural Actor-Critic (NAC). A number of different approaches have been taken to approximate the policy Hessian, but none delivered a proper Hessian estimate. Therefore, no LM actor update method was created. In VGBP the action is found by optimization of the right hand side of the Bellman equation. VGBP uses the provided reward function and a process model for this optimization. The process model is learned online using local linear regression (LLR). Due to the efficient use of information VGBP shows fast learning on the pendulum and a 2-DOF robotic arm.

Systems and Control

Delft Center for Systems and Control

Mechanical, Maritime and Materials Engineering

Advisors/Committee Members: Grondman, I. (mentor), Babuska, R. (mentor).

Subjects/Keywords: reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Van Rooijen, J. C. (. (2012). Learning Parameter Selection in Continuous Reinforcement Learning: Attempting to Reduce Tuning Effords. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:94b81bc2-aff6-457f-9b54-be5e005def38

Chicago Manual of Style (16^th Edition):

Van Rooijen, J C (author). “Learning Parameter Selection in Continuous Reinforcement Learning: Attempting to Reduce Tuning Effords.” 2012. Masters Thesis, Delft University of Technology. Accessed March 04, 2021. http://resolver.tudelft.nl/uuid:94b81bc2-aff6-457f-9b54-be5e005def38.

MLA Handbook (7^th Edition):

Van Rooijen, J C (author). “Learning Parameter Selection in Continuous Reinforcement Learning: Attempting to Reduce Tuning Effords.” 2012. Web. 04 Mar 2021.

Vancouver:

Van Rooijen JC(. Learning Parameter Selection in Continuous Reinforcement Learning: Attempting to Reduce Tuning Effords. [Internet] [Masters thesis]. Delft University of Technology; 2012. [cited 2021 Mar 04]. Available from: http://resolver.tudelft.nl/uuid:94b81bc2-aff6-457f-9b54-be5e005def38.

Council of Science Editors:

Van Rooijen JC(. Learning Parameter Selection in Continuous Reinforcement Learning: Attempting to Reduce Tuning Effords. [Masters Thesis]. Delft University of Technology; 2012. Available from: http://resolver.tudelft.nl/uuid:94b81bc2-aff6-457f-9b54-be5e005def38

Delft University of Technology

7. Van Diepen, M.D.M. (author). Avoiding failure states during reinforcement learning.

Degree: 2011, Delft University of Technology

URL: http://resolver.tudelft.nl/uuid:1f03c580-9fd5-4807-87b5-d70890e05ff6

►

The Delft Biorobotics Laboratory develops bipedal humanoid robots. One of these robots, called LEO, is designed to learn to walk using reinforcement learning. During learning,… (more)

▼

The Delft Biorobotics Laboratory develops bipedal humanoid robots. One of these robots, called LEO, is designed to learn to walk using reinforcement learning. During learning, LEO will make mistakes and fall. These mistakes can cause serious dam- age to the system but are an integral part of the learning process. A likely solution is punishing the robot more severely for falling. However, punishing the robot too rigorously can lead to a robot that is too cautious to make a step. In this research, three methods were tested that reduce the amount of falls during learning, without restricting the possible solutions or increasing the learning time. We introduce Threshold Restricted Learning (TRL), a new method for action se- lection that, during exploration, makes the probability an action is chosen dependent on the expected reward for taking that action. Actions with expected rewards below the set threshold have a signifcant reduced probability of being chosen. The concept of TRL developed from the desire to optimize the use of a pre-learned solution. Hence, TRL was tested after pre-learning in simulation. The largest reduction in the amount of falls achieved by TRL in this research was 50% TRL did not increase the learning time. Without pre-learning the largest reduction was 7%. Softmax action selection is a well known but underused selection method. Combined with pre-learning it was able to reduce the amount of falls during learning with approximately 80% and did not increase the learning time either. Without pre-learning, Softmax could still achieve a reduction of approximately 20%. Sarsa2Q, another new method, stores expected rewards at different levels op generalization. The level with little generalization is used to store the expected total positive reward, while the level with more generalization is used to store the expected total negative reward. Sarsa2Q learned to avoid the failure states for the inverted pendulum problem faster than using a single level of generalization. This method does not use pre-learning. The highest achieved reduction in the amount of falls was approximately 20%. Sarsa2Q can have a broader use than just avoiding failure states. Rather than only using the coarse generalization for negative rewards, it can be used for all rewards that need less precision.

Biorobotics

BioMechanical Engineering

Mechanical, Maritime and Materials Engineering

Advisors/Committee Members: Jonker, P.P. (mentor), Schuitema, E. (mentor).

Subjects/Keywords: reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Van Diepen, M. D. M. (. (2011). Avoiding failure states during reinforcement learning. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:1f03c580-9fd5-4807-87b5-d70890e05ff6

Chicago Manual of Style (16^th Edition):

Van Diepen, M D M (author). “Avoiding failure states during reinforcement learning.” 2011. Masters Thesis, Delft University of Technology. Accessed March 04, 2021. http://resolver.tudelft.nl/uuid:1f03c580-9fd5-4807-87b5-d70890e05ff6.

MLA Handbook (7^th Edition):

Van Diepen, M D M (author). “Avoiding failure states during reinforcement learning.” 2011. Web. 04 Mar 2021.

Vancouver:

Van Diepen MDM(. Avoiding failure states during reinforcement learning. [Internet] [Masters thesis]. Delft University of Technology; 2011. [cited 2021 Mar 04]. Available from: http://resolver.tudelft.nl/uuid:1f03c580-9fd5-4807-87b5-d70890e05ff6.

Council of Science Editors:

Van Diepen MDM(. Avoiding failure states during reinforcement learning. [Masters Thesis]. Delft University of Technology; 2011. Available from: http://resolver.tudelft.nl/uuid:1f03c580-9fd5-4807-87b5-d70890e05ff6

University of Illinois – Urbana-Champaign

8. Potok, Matthew. Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study.

Degree: MS, Electrical & Computer Engr, 2018, University of Illinois – Urbana-Champaign

URL: http://hdl.handle.net/2142/102518

► Reinforcement learning (RL) is a general method for agents to learn optimal control policies through exploration and experience. Due to its generality, RL can generate… (more)

Subjects/Keywords: Reinforcement Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Potok, M. (2018). Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study. (Thesis). University of Illinois – Urbana-Champaign. Retrieved from http://hdl.handle.net/2142/102518

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Potok, Matthew. “Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study.” 2018. Thesis, University of Illinois – Urbana-Champaign. Accessed March 04, 2021. http://hdl.handle.net/2142/102518.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Potok, Matthew. “Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study.” 2018. Web. 04 Mar 2021.

Vancouver:

Potok M. Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study. [Internet] [Thesis]. University of Illinois – Urbana-Champaign; 2018. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/2142/102518.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Potok M. Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study. [Thesis]. University of Illinois – Urbana-Champaign; 2018. Available from: http://hdl.handle.net/2142/102518

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Delft University of Technology

9. Keulen, Bart (author). Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning.

Degree: 2018, Delft University of Technology

URL: http://resolver.tudelft.nl/uuid:eca20454-7905-42a1-9fb0-f72776fd5422

► An important problem in reinforcement learning is the exploration-exploitation dilemma. Especially for environments with sparse or misleading rewards it has proven difficult to construct a… (more)

▼ An important problem in reinforcement learning is the exploration-exploitation dilemma. Especially for environments with sparse or misleading rewards it has proven difficult to construct a good exploration strategy. For discrete domains good exploration strategies have been devised, but are often nontrivial to implement on more complex domains with continuous states and/or actions. In this work, a novel persistent and directed exploration framework is developed, called Smart Start. Usually, a reinforcement learning agent executes its learned policy with some exploration strategy from the start until the end of an episode, which we call ``normal'' learning. The idea of Smart Start is to split a reinforcement learning episode in two parts, the Smart Start phase and the ``normal'' learning phase. The initial Smart Start phase guides the agent to a region in which the agent expects to learn the most. The region is constructed using previous experiences and the guiding is done using a model-based planning or trajectory optimization method. When the agent arrives at the region, it continues its ``normal'' reinforcement learning. This approach leaves the performance of the used reinforcement learning algorithm unchanged, but augments it with persistent and directed exploration. The Smart Start framework was evaluated using three reinforcement learning algorithms, a simple model-based reinforcement learning algorithm (MBRL), R-max and Q-Learning with epsilon-greedy, Boltzmann and UCB1 exploration. The evaluation was done on four discrete gridworld environments. Three environments with sparse rewards and one with misleading rewards. We showed that the performance of Q-Learning with Smart Start is comparable to R-max, which performs near optimal in the used scenarios. The MBRL algorithm with Smart Start is even able to outperform R-max in some of the problems. We show that Smart Start is a good framework for exploration that can be incorporated with any reinforcement learning algorithm. Advisors/Committee Members: Kober, Jens (mentor), Pratt, Jerry (graduation committee), Delft University of Technology (degree granting institution).

Subjects/Keywords: Reinforcement Learning; Exploration

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Keulen, B. (. (2018). Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:eca20454-7905-42a1-9fb0-f72776fd5422

Chicago Manual of Style (16^th Edition):

Keulen, Bart (author). “Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning.” 2018. Masters Thesis, Delft University of Technology. Accessed March 04, 2021. http://resolver.tudelft.nl/uuid:eca20454-7905-42a1-9fb0-f72776fd5422.

MLA Handbook (7^th Edition):

Keulen, Bart (author). “Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning.” 2018. Web. 04 Mar 2021.

Vancouver:

Keulen B(. Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning. [Internet] [Masters thesis]. Delft University of Technology; 2018. [cited 2021 Mar 04]. Available from: http://resolver.tudelft.nl/uuid:eca20454-7905-42a1-9fb0-f72776fd5422.

Council of Science Editors:

Keulen B(. Smart Start: A Directed and Persistent Exploration Framework for Reinforcement Learning. [Masters Thesis]. Delft University of Technology; 2018. Available from: http://resolver.tudelft.nl/uuid:eca20454-7905-42a1-9fb0-f72776fd5422

University of Illinois – Urbana-Champaign

10. Van Stralen, Neale A. Hierarchical reinforcement learning for adaptive and autonomous decision-making in robotics.

Degree: MS, Aerospace Engineering, 2020, University of Illinois – Urbana-Champaign

URL: http://hdl.handle.net/2142/108536

► In recent years, Reinforcement Learning has been able to solve extremely complex games in simulation, but with limited success in deployment to real-world scenarios. The… (more)

Subjects/Keywords: Reinforcement Learning; Adaptation

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Van Stralen, N. A. (2020). Hierarchical reinforcement learning for adaptive and autonomous decision-making in robotics. (Thesis). University of Illinois – Urbana-Champaign. Retrieved from http://hdl.handle.net/2142/108536

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Van Stralen, Neale A. “Hierarchical reinforcement learning for adaptive and autonomous decision-making in robotics.” 2020. Thesis, University of Illinois – Urbana-Champaign. Accessed March 04, 2021. http://hdl.handle.net/2142/108536.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Van Stralen, Neale A. “Hierarchical reinforcement learning for adaptive and autonomous decision-making in robotics.” 2020. Web. 04 Mar 2021.

Vancouver:

Van Stralen NA. Hierarchical reinforcement learning for adaptive and autonomous decision-making in robotics. [Internet] [Thesis]. University of Illinois – Urbana-Champaign; 2020. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/2142/108536.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Van Stralen NA. Hierarchical reinforcement learning for adaptive and autonomous decision-making in robotics. [Thesis]. University of Illinois – Urbana-Champaign; 2020. Available from: http://hdl.handle.net/2142/108536

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

University of New South Wales

11. Ismail, Hafsa. A neural network framework for combining different task types and motivations in motivated reinforcement learning.

Degree: Engineering & Information Technology, 2014, University of New South Wales

URL: http://handle.unsw.edu.au/1959.4/53975 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:12686/SOURCE02?view=true

► Combining different motivation models for different task types within artificial agents has the potential to produce agents capable of a greater range of behaviours in… (more)

▼ Combining different motivation models for different task types within artificial agents has the potential to produce agents capable of a greater range of behaviours in different situations. This thesis will explore the ability to produce agents that can identify different types of tasks online then learn solutions for these tasks while driven by one or more motivations in isolation or combination to produce different agent behaviour that will focus on different task types. The main contributions of this research are: Definitions for four task types for use in motivated reinforcement learning agents; a multi-layer neural network model for combining different motivations for different task types; a study of agent ability to identify different types of tasks in a simulated game scenario. A review of existing computational models of motivation and the layered approaches to the design of goal generation and intrinsically motivated learning agents inspire the multi-layer neural network framework presented and examined experimentally in this thesis. The first set of experiments examines topological and non-topological adaptive resonance theory networks that are used to classify sensed states and identify different types of tasks. Different distance metrics and weight update rules are used to determine the impact of the network structure on identified task types. The most promising network structure, distance metric and update rule is then chosen to implement the task identification process in conjunction with motivated reinforcement learning. In the second set of experiments, three motivation models, novelty, interest and competence, are examined separately in conjunction with our task identification algorithm. Established metrics and statistical methods are used to quantify learning behaviour and the complexity and variety of repetitive behavioural structures. Finally, we examine agent behaviour when the agent can identify more than one type of task with a single motive model that drives the learning process. Parallel and series task types combination modes are explored. Results show that a series approach to changing the type of task the agent can identify is most effective for producing motivated reinforcement learning agents that can both learn effectively and exhibit quantifiable changes in their characteristic learning behaviour over time. Advisors/Committee Members: Merrick, Kathryn, Engineering & Information Technology, Australian Defence Force Academy, UNSW, Barlow, Michael, Engineering & Information Technology, Australian Defence Force Academy, UNSW.

Subjects/Keywords: Motivated Reinforcement Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Ismail, H. (2014). A neural network framework for combining different task types and motivations in motivated reinforcement learning. (Masters Thesis). University of New South Wales. Retrieved from http://handle.unsw.edu.au/1959.4/53975 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:12686/SOURCE02?view=true

Chicago Manual of Style (16^th Edition):

Ismail, Hafsa. “A neural network framework for combining different task types and motivations in motivated reinforcement learning.” 2014. Masters Thesis, University of New South Wales. Accessed March 04, 2021. http://handle.unsw.edu.au/1959.4/53975 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:12686/SOURCE02?view=true.

MLA Handbook (7^th Edition):

Ismail, Hafsa. “A neural network framework for combining different task types and motivations in motivated reinforcement learning.” 2014. Web. 04 Mar 2021.

Vancouver:

Ismail H. A neural network framework for combining different task types and motivations in motivated reinforcement learning. [Internet] [Masters thesis]. University of New South Wales; 2014. [cited 2021 Mar 04]. Available from: http://handle.unsw.edu.au/1959.4/53975 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:12686/SOURCE02?view=true.

Council of Science Editors:

Ismail H. A neural network framework for combining different task types and motivations in motivated reinforcement learning. [Masters Thesis]. University of New South Wales; 2014. Available from: http://handle.unsw.edu.au/1959.4/53975 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:12686/SOURCE02?view=true

Oregon State University

12. Wilson, Aaron (Aaron Creighton). Bayesian methods for knowledge transfer and policy search in reinforcement learning.

Degree: PhD, Computer Science, 2012, Oregon State University

URL: http://hdl.handle.net/1957/34550

► How can an agent generalize its knowledge to new circumstances? To learn effectively an agent acting in a sequential decision problem must make intelligent action… (more)

▼ How can an agent generalize its knowledge to new circumstances? To learn effectively an agent acting in a sequential decision problem must make intelligent action selection choices based on its available knowledge. This dissertation focuses on Bayesian methods of representing learned knowledge and develops novel algorithms that exploit the represented knowledge when selecting actions. Our first contribution introduces the multi-task Reinforcement Learning setting in which an agent solves a sequence of tasks. An agent equipped with knowledge of the relationship between tasks can transfer knowledge between them. We propose the transfer of two distinct types of knowledge: knowledge of domain models and knowledge of policies. To represent the transferable knowledge, we propose hierarchical Bayesian priors on domain models and policies respectively. To transfer domain model knowledge, we introduce a new algorithm for model-based Bayesian Reinforcement Learning in the multi-task setting which exploits the learned hierarchical Bayesian model to improve exploration in related tasks. To transfer policy knowledge, we introduce a new policy search algorithm that accepts a policy prior as input and uses the prior to bias policy search. A specific implementation of this algorithm is developed that accepts a hierarchical policy prior. The algorithm learns the hierarchical structure and reuses components of the structure in related tasks. Our second contribution addresses the basic problem of generalizing knowledge gained from previously-executed policies. Bayesian Optimization is a method of exploiting a prior model of an objective function to quickly identify the point maximizing the modeled objective. Successful use of Bayesian Optimization in Reinforcement Learning requires a model relating policies and their performance. Given such a model, Bayesian Optimization can be applied to search for an optimal policy. Early work using Bayesian Optimization in the Reinforcement Learning setting ignored the sequential nature of the underlying decision problem. The work presented in this thesis explicitly addresses this problem. We construct new Bayesian models that take advantage of sequence information to better generalize knowledge across policies. We empirically evaluate the value of this approach in a variety of Reinforcement Learning benchmark problems. Experiments show that our method significantly reduces the amount of exploration required to identify the optimal policy. Our final contribution is a new framework for learning parametric policies from queries presented to an expert. In many domains it is difficult to provide expert demonstrations of desired policies. However, it may still be a simple matter for an expert to identify good and bad performance. To take advantage of this limited expert knowledge, our agent presents experts with pairs of demonstrations and asks which of the demonstrations best represents a latent target behavior. The goal is to use a small number of… Advisors/Committee Members: Tadepalli, Prasad (advisor), Tom, Dietterich (committee member).

Subjects/Keywords: Machine Learning; Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Wilson, A. (. C. (2012). Bayesian methods for knowledge transfer and policy search in reinforcement learning. (Doctoral Dissertation). Oregon State University. Retrieved from http://hdl.handle.net/1957/34550

Chicago Manual of Style (16^th Edition):

Wilson, Aaron (Aaron Creighton). “Bayesian methods for knowledge transfer and policy search in reinforcement learning.” 2012. Doctoral Dissertation, Oregon State University. Accessed March 04, 2021. http://hdl.handle.net/1957/34550.

MLA Handbook (7^th Edition):

Wilson, Aaron (Aaron Creighton). “Bayesian methods for knowledge transfer and policy search in reinforcement learning.” 2012. Web. 04 Mar 2021.

Vancouver:

Wilson A(C. Bayesian methods for knowledge transfer and policy search in reinforcement learning. [Internet] [Doctoral dissertation]. Oregon State University; 2012. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/1957/34550.

Council of Science Editors:

Wilson A(C. Bayesian methods for knowledge transfer and policy search in reinforcement learning. [Doctoral Dissertation]. Oregon State University; 2012. Available from: http://hdl.handle.net/1957/34550

Rutgers University

13. Marivate, Vukosi N. Improved empirical methods in reinforcement-learning evaluation.

Degree: PhD, Computer Science, 2015, Rutgers University

URL: https://rucore.libraries.rutgers.edu/rutgers-lib/46389/

►

The central question addressed in this research is ”can we define evaluation methodologies that encourage reinforcement-learning (RL) algorithms to work effectively with real-life data?” First,… (more)

Subjects/Keywords: Reinforcement learning; Machine learning; Algorithms

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Marivate, V. N. (2015). Improved empirical methods in reinforcement-learning evaluation. (Doctoral Dissertation). Rutgers University. Retrieved from https://rucore.libraries.rutgers.edu/rutgers-lib/46389/

Chicago Manual of Style (16^th Edition):

Marivate, Vukosi N. “Improved empirical methods in reinforcement-learning evaluation.” 2015. Doctoral Dissertation, Rutgers University. Accessed March 04, 2021. https://rucore.libraries.rutgers.edu/rutgers-lib/46389/.

MLA Handbook (7^th Edition):

Marivate, Vukosi N. “Improved empirical methods in reinforcement-learning evaluation.” 2015. Web. 04 Mar 2021.

Vancouver:

Marivate VN. Improved empirical methods in reinforcement-learning evaluation. [Internet] [Doctoral dissertation]. Rutgers University; 2015. [cited 2021 Mar 04]. Available from: https://rucore.libraries.rutgers.edu/rutgers-lib/46389/.

Council of Science Editors:

Marivate VN. Improved empirical methods in reinforcement-learning evaluation. [Doctoral Dissertation]. Rutgers University; 2015. Available from: https://rucore.libraries.rutgers.edu/rutgers-lib/46389/

University of Aberdeen

14. Alexander, John W. Transfer in reinforcement learning.

Degree: PhD, 2015, University of Aberdeen

URL: https://abdn.alma.exlibrisgroup.com/view/delivery/44ABE_INST/12152715040005941 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.675561

► The problem of developing skill repertoires autonomously in robotics and artificial intelligence is becoming ever more pressing. Currently, the issues of how to apply prior… (more)

▼ The problem of developing skill repertoires autonomously in robotics and artificial intelligence is becoming ever more pressing. Currently, the issues of how to apply prior knowledge to new situations and which knowledge to apply have not been sufficiently studied. We present a transfer setting where a reinforcement learning agent faces multiple problem solving tasks drawn from an unknown generative process, where each task has similar dynamics. The task dynamics are changed by varying in the transition function between states. The tasks are presented sequentially with the latest task presented considered as the target for transfer. We describe two approaches to solving this problem. Firstly we present an algorithm for transfer of the function encoding the stateaction value, defined as value function transfer. This algorithm uses the value function of a source policy to initialise the policy of a target task. We varied the type of basis the algorithm used to approximate the value function. Empirical results in several well known domains showed that the learners benefited from the transfer in the majority of cases. Results also showed that the Radial basis performed better in general than the Fourier. However contrary to expectation the Fourier basis benefited most from the transfer. Secondly, we present an algorithm for learning an informative prior which encodes beliefs about the underlying dynamics shared across all tasks. We call this agent the Informative Prior agent (IP). The prior is learnt though experience and captures the commonalities in the transition dynamics of the domain and allows for a quantification of the agent's uncertainty about these. By using a sparse distribution of the uncertainty in the dynamics as a prior, the IP agent can successfully learn a model of 1) the set of feasible transitions rather than the set of possible transitions, and 2) the likelihood of each of the feasible transitions. Analysis focusing on the accuracy of the learned model showed that IP had a very good accuracy bound, which is expressible in terms of only the permissible error and the diffusion, a factor that describes the concentration of the prior mass around the truth, and which decreases as the number of tasks experienced grows. The empirical evaluation of IP showed that an agent which uses the informative prior outperforms several existing Bayesian reinforcement learning algorithms on tasks with shared structure in a domain where multiple related tasks were presented only once to the learners. IP is a step towards the autonomous acquisition of behaviours in artificial intelligence. IP also provides a contribution towards the analysis of exploration and exploitation in the transfer paradigm.

Subjects/Keywords: 004; Reinforcement learning; Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Alexander, J. W. (2015). Transfer in reinforcement learning. (Doctoral Dissertation). University of Aberdeen. Retrieved from https://abdn.alma.exlibrisgroup.com/view/delivery/44ABE_INST/12152715040005941 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.675561

Chicago Manual of Style (16^th Edition):

Alexander, John W. “Transfer in reinforcement learning.” 2015. Doctoral Dissertation, University of Aberdeen. Accessed March 04, 2021. https://abdn.alma.exlibrisgroup.com/view/delivery/44ABE_INST/12152715040005941 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.675561.

MLA Handbook (7^th Edition):

Alexander, John W. “Transfer in reinforcement learning.” 2015. Web. 04 Mar 2021.

Vancouver:

Alexander JW. Transfer in reinforcement learning. [Internet] [Doctoral dissertation]. University of Aberdeen; 2015. [cited 2021 Mar 04]. Available from: https://abdn.alma.exlibrisgroup.com/view/delivery/44ABE_INST/12152715040005941 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.675561.

Council of Science Editors:

Alexander JW. Transfer in reinforcement learning. [Doctoral Dissertation]. University of Aberdeen; 2015. Available from: https://abdn.alma.exlibrisgroup.com/view/delivery/44ABE_INST/12152715040005941 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.675561

University of Waterloo

15. Gaurav, Ashish. Safety-Oriented Stability Biases for Continual Learning.

Degree: 2020, University of Waterloo

URL: http://hdl.handle.net/10012/15579

► Continual learning is often confounded by “catastrophic forgetting” that prevents neural networks from learning tasks sequentially. In the case of real world classification systems that… (more)

Subjects/Keywords: deep learning; continual learning; classification; reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Gaurav, A. (2020). Safety-Oriented Stability Biases for Continual Learning. (Thesis). University of Waterloo. Retrieved from http://hdl.handle.net/10012/15579

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Gaurav, Ashish. “Safety-Oriented Stability Biases for Continual Learning.” 2020. Thesis, University of Waterloo. Accessed March 04, 2021. http://hdl.handle.net/10012/15579.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Gaurav, Ashish. “Safety-Oriented Stability Biases for Continual Learning.” 2020. Web. 04 Mar 2021.

Vancouver:

Gaurav A. Safety-Oriented Stability Biases for Continual Learning. [Internet] [Thesis]. University of Waterloo; 2020. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10012/15579.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Gaurav A. Safety-Oriented Stability Biases for Continual Learning. [Thesis]. University of Waterloo; 2020. Available from: http://hdl.handle.net/10012/15579

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Hong Kong University of Science and Technology

16. Wellmer, Zachary William CSE. Building and leveraging implicit models for policy gradient methods.

Degree: 2019, Hong Kong University of Science and Technology

URL: http://repository.ust.hk/ir/Record/1783.1-102375 ; https://doi.org/10.14711/thesis-991012757568703412 ; http://repository.ust.hk/ir/bitstream/1783.1-102375/1/th_redirect.html

► In this thesis, we study Policy Prediction Network and Policy Tree Network, both are deep reinforcement learning architectures offering ways to improve sample complexity and… (more)

Subjects/Keywords: Reinforcement learning ; Machine learning ; Implicit learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Wellmer, Z. W. C. (2019). Building and leveraging implicit models for policy gradient methods. (Thesis). Hong Kong University of Science and Technology. Retrieved from http://repository.ust.hk/ir/Record/1783.1-102375 ; https://doi.org/10.14711/thesis-991012757568703412 ; http://repository.ust.hk/ir/bitstream/1783.1-102375/1/th_redirect.html

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Wellmer, Zachary William CSE. “Building and leveraging implicit models for policy gradient methods.” 2019. Thesis, Hong Kong University of Science and Technology. Accessed March 04, 2021. http://repository.ust.hk/ir/Record/1783.1-102375 ; https://doi.org/10.14711/thesis-991012757568703412 ; http://repository.ust.hk/ir/bitstream/1783.1-102375/1/th_redirect.html.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Wellmer, Zachary William CSE. “Building and leveraging implicit models for policy gradient methods.” 2019. Web. 04 Mar 2021.

Vancouver:

Wellmer ZWC. Building and leveraging implicit models for policy gradient methods. [Internet] [Thesis]. Hong Kong University of Science and Technology; 2019. [cited 2021 Mar 04]. Available from: http://repository.ust.hk/ir/Record/1783.1-102375 ; https://doi.org/10.14711/thesis-991012757568703412 ; http://repository.ust.hk/ir/bitstream/1783.1-102375/1/th_redirect.html.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Wellmer ZWC. Building and leveraging implicit models for policy gradient methods. [Thesis]. Hong Kong University of Science and Technology; 2019. Available from: http://repository.ust.hk/ir/Record/1783.1-102375 ; https://doi.org/10.14711/thesis-991012757568703412 ; http://repository.ust.hk/ir/bitstream/1783.1-102375/1/th_redirect.html

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

University of Texas – Austin

17. Shah, Rishi Alpesh. Deep R learning for continual area sweeping.

Degree: MSin Computational Science, Engineering, and Mathematics, Computational Science, Engineering, and Mathematics, 2019, University of Texas – Austin

URL: http://dx.doi.org/10.26153/tsw/8201

► In order to maintain robustness, autonomous robots need to constantly update their knowledge of the environment, which can be expensive when they are deployed in… (more)

Subjects/Keywords: Machine learning; Reinforcement learning; Robotics; Robot learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Shah, R. A. (2019). Deep R learning for continual area sweeping. (Masters Thesis). University of Texas – Austin. Retrieved from http://dx.doi.org/10.26153/tsw/8201

Chicago Manual of Style (16^th Edition):

Shah, Rishi Alpesh. “Deep R learning for continual area sweeping.” 2019. Masters Thesis, University of Texas – Austin. Accessed March 04, 2021. http://dx.doi.org/10.26153/tsw/8201.

MLA Handbook (7^th Edition):

Shah, Rishi Alpesh. “Deep R learning for continual area sweeping.” 2019. Web. 04 Mar 2021.

Vancouver:

Shah RA. Deep R learning for continual area sweeping. [Internet] [Masters thesis]. University of Texas – Austin; 2019. [cited 2021 Mar 04]. Available from: http://dx.doi.org/10.26153/tsw/8201.

Council of Science Editors:

Shah RA. Deep R learning for continual area sweeping. [Masters Thesis]. University of Texas – Austin; 2019. Available from: http://dx.doi.org/10.26153/tsw/8201

University of Illinois – Urbana-Champaign

18. Benson, Christopher Edward. Improving cache replacement policy using deep reinforcement learning.

Degree: MS, Computer Science, 2018, University of Illinois – Urbana-Champaign

URL: http://hdl.handle.net/2142/102858

► This thesis explores the use of reinforcement learning approaches to improve replacement policies of caches. In today's internet, caches play a vital role in improving… (more)

Subjects/Keywords: Reinforcement Learning; Machine Learning; Deep Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Benson, C. E. (2018). Improving cache replacement policy using deep reinforcement learning. (Thesis). University of Illinois – Urbana-Champaign. Retrieved from http://hdl.handle.net/2142/102858

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Benson, Christopher Edward. “Improving cache replacement policy using deep reinforcement learning.” 2018. Thesis, University of Illinois – Urbana-Champaign. Accessed March 04, 2021. http://hdl.handle.net/2142/102858.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Benson, Christopher Edward. “Improving cache replacement policy using deep reinforcement learning.” 2018. Web. 04 Mar 2021.

Vancouver:

Benson CE. Improving cache replacement policy using deep reinforcement learning. [Internet] [Thesis]. University of Illinois – Urbana-Champaign; 2018. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/2142/102858.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Benson CE. Improving cache replacement policy using deep reinforcement learning. [Thesis]. University of Illinois – Urbana-Champaign; 2018. Available from: http://hdl.handle.net/2142/102858

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Delft University of Technology

19. Kovács, B. (author). Safe reinforcement learning in long-horizon partially observable environments.

Degree: 2020, Delft University of Technology

URL: http://resolver.tudelft.nl/uuid:83556b37-1484-41f4-b168-3bd316def4a0

► Deep reinforcement learning went through an unprecedented development in the last decade, resulting in agents defeating world champion human players in complex board games like… (more)

▼ Deep reinforcement learning went through an unprecedented development in the last decade, resulting in agents defeating world champion human players in complex board games like go and chess. With few exceptions, deep reinforcement learning research focuses on fully observable environments, while there is slightly less research in the direction of partially observable environments. Though, they are relevant in real-world applications where the physical systems' sensors can read only a limited subset of features required for decision making. In many problems, long-term memory is required for the agent to make optimal decisions, possibly through tens or hundreds of timesteps. In supervised deep learning, in the NLP domain, sequential input processing is done with specialized network architectures like variants of RNNs and attention-based networks. In this work, we investigate the potential of these advanced sequence-processing architectures in the context of deep reinforcement learning for partially observable environments. Additionally, since partial observability widely appears in the physical world, we take a safe approach by trying to limit high exploration costs and damage to the agent and its environment. First, we augment the soft actor-critic method with constraints on the episodic cost, resulting in an objective function with two Lagrangian multiplier: an entropy temperature and a safety temperature. Then, to support long-horizon, partially observable environments, we use gated recurrent (LSTM, GRU) and self-attention based neural networks for the policy and the estimation of Q-functions. We also study how the design choices and hyperparameters of the self-attention based method affect the performance. To evaluate the problem in safety-constrained environments with long-term temporal dependencies, we develop a new set of benchmarks with four parameterizable, partially observable simulations. The environments are also parameterizable with the length of the history containing relevant features. Hence, we can observe how different network architectures handle the same problems with varying time horizons. Additionally, we introduce a practical framework for the reproducible evaluation of the methods. We conclude that both the recurrent and the self-attention based architectures have high application potential in the introduced domains. We confirm that the feedforward network based baseline agent, shows high performance on problems where only a few, or tens of timesteps have to be processed sequentially. The recurrent and self-attention based architectures show their advantage in environments with longer horizons, where the sequence of events play an important role and looking back to a fixed position is not sufficient. Maintaining safety proves to be problematic when the reward and cost functions show correlation. Additionally, strict cost limits usually lead to a poor policy and no exploration, contributing to higher costs on the long term, for some environments. We… Advisors/Committee Members: Spaan, M.T.J. (mentor), Yang, Q. (graduation committee), van Gemert, J.C. (graduation committee), Kober, J. (graduation committee), Delft University of Technology (degree granting institution).

Subjects/Keywords: reinforcement learning; self-attention; partially observable; deep reinforcement learning; safe reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Kovács, B. (. (2020). Safe reinforcement learning in long-horizon partially observable environments. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:83556b37-1484-41f4-b168-3bd316def4a0

Chicago Manual of Style (16^th Edition):

Kovács, B (author). “Safe reinforcement learning in long-horizon partially observable environments.” 2020. Masters Thesis, Delft University of Technology. Accessed March 04, 2021. http://resolver.tudelft.nl/uuid:83556b37-1484-41f4-b168-3bd316def4a0.

MLA Handbook (7^th Edition):

Kovács, B (author). “Safe reinforcement learning in long-horizon partially observable environments.” 2020. Web. 04 Mar 2021.

Vancouver:

Kovács B(. Safe reinforcement learning in long-horizon partially observable environments. [Internet] [Masters thesis]. Delft University of Technology; 2020. [cited 2021 Mar 04]. Available from: http://resolver.tudelft.nl/uuid:83556b37-1484-41f4-b168-3bd316def4a0.

Council of Science Editors:

Kovács B(. Safe reinforcement learning in long-horizon partially observable environments. [Masters Thesis]. Delft University of Technology; 2020. Available from: http://resolver.tudelft.nl/uuid:83556b37-1484-41f4-b168-3bd316def4a0

NSYSU

20. Lin, Kun-da. Deep Reinforcement Learning with a Gating Network.

Degree: Master, Electrical Engineering, 2017, NSYSU

URL: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0223117-131536

► Reinforcement Learning (RL) is a good way to train the robot since it doesn't need an exact model of the environment. All need is to… (more)

▼ Reinforcement Learning (RL) is a good way to train the robot since it doesn't need an exact model of the environment. All need is to let a learning agent interact with the en-vironment by an appropriate reward function, which is associated with the goal of the task that the agent is expected to accomplish. Unfortunately, itâs hard to learn a diffi-cult reward function for a complicated problem, such as a soccer player in the game where the goal of scoring is not directly related to the mission or the role the player is asked to play by the coach. Besides, the tabular method for approximation of returns in RL is more suitable for an environment with less states. In a huge state space, RL methods always face the curse of dimensionality. To alleviate those difficulties, this paper proposed an algorithm â a deep reinforcement learning method regulated by and gating networks. By the merit of deep learning neural networks, even regarding pixels in an image as states, the latent features can be trained and implicitly extracted layer by layer from raw data. In the proposed method, a composed policy can be obtained by a gating network which regulates the outputs from several deep learning modules, each of which is trained for an individual policy. In this thesis, two video games, flappy bird, and ping-pong is adopted as the testbeds to examinate the performance of the proposed method. In the proposed architecture, each policy module of deep learning is trained by a simple reward functions first. By the gating networks, these simple policies can be composed into a more sophisticated one, so as to accommodate with more complicated tasks. This is akin to the divide-and-conquer strategy. The proposed architecture has two kinds of arrangements and structures. One is called the in-parallel gating network, and the other is called the in-serial network. From the outcomes, itâs observed that both can efficiently shorten the training time. Advisors/Committee Members: Ming-Yi Ju (chair), Yu-Jen Chen (chair), Kao-Shing Hwang (committee member), Jin-Ling Lin (chair).

Subjects/Keywords: Reinforcement Learning; Deep Reinforcement Learning; Deep Learning; Gating network; Neural network

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Lin, K. (2017). Deep Reinforcement Learning with a Gating Network. (Thesis). NSYSU. Retrieved from http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0223117-131536

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Lin, Kun-da. “Deep Reinforcement Learning with a Gating Network.” 2017. Thesis, NSYSU. Accessed March 04, 2021. http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0223117-131536.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Lin, Kun-da. “Deep Reinforcement Learning with a Gating Network.” 2017. Web. 04 Mar 2021.

Vancouver:

Lin K. Deep Reinforcement Learning with a Gating Network. [Internet] [Thesis]. NSYSU; 2017. [cited 2021 Mar 04]. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0223117-131536.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Lin K. Deep Reinforcement Learning with a Gating Network. [Thesis]. NSYSU; 2017. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0223117-131536

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

NSYSU

21. Tseng, Yi-Chia. An Unified Approach to Inverse Reinforcement Learning by Oppositive Demonstrations.

Degree: Master, Electrical Engineering, 2015, NSYSU

URL: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0727115-130716

► Reinforcement learning (RL) techniques use a reward function to correct a learning agent to solve sequential decision making problems through interactions with a dynamic environment,… (more)

Subjects/Keywords: Apprenticeship Learning; Feature weight; Inverse Reinforcement learning; Reward function; Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Tseng, Y. (2015). An Unified Approach to Inverse Reinforcement Learning by Oppositive Demonstrations. (Thesis). NSYSU. Retrieved from http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0727115-130716

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Tseng, Yi-Chia. “An Unified Approach to Inverse Reinforcement Learning by Oppositive Demonstrations.” 2015. Thesis, NSYSU. Accessed March 04, 2021. http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0727115-130716.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Tseng, Yi-Chia. “An Unified Approach to Inverse Reinforcement Learning by Oppositive Demonstrations.” 2015. Web. 04 Mar 2021.

Vancouver:

Tseng Y. An Unified Approach to Inverse Reinforcement Learning by Oppositive Demonstrations. [Internet] [Thesis]. NSYSU; 2015. [cited 2021 Mar 04]. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0727115-130716.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Tseng Y. An Unified Approach to Inverse Reinforcement Learning by Oppositive Demonstrations. [Thesis]. NSYSU; 2015. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0727115-130716

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

NSYSU

22. Lin, Hung-shyuan. Applying The Concept of Fuzzy Logic to Inverse Reinforcement Learning.

Degree: Master, Electrical Engineering, 2015, NSYSU

URL: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-1025115-185021

► Itâs a study on Reinforcement Learning, learning interaction of agents and dynamic environment to get reward function R, and update the policy, converge learning and… (more)

Subjects/Keywords: Inverse reinforcement learning; Reward function; Fuzzy; Reinforcement learning; AdaBoost; Apprenticeship learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Lin, H. (2015). Applying The Concept of Fuzzy Logic to Inverse Reinforcement Learning. (Thesis). NSYSU. Retrieved from http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-1025115-185021

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Lin, Hung-shyuan. “Applying The Concept of Fuzzy Logic to Inverse Reinforcement Learning.” 2015. Thesis, NSYSU. Accessed March 04, 2021. http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-1025115-185021.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Lin, Hung-shyuan. “Applying The Concept of Fuzzy Logic to Inverse Reinforcement Learning.” 2015. Web. 04 Mar 2021.

Vancouver:

Lin H. Applying The Concept of Fuzzy Logic to Inverse Reinforcement Learning. [Internet] [Thesis]. NSYSU; 2015. [cited 2021 Mar 04]. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-1025115-185021.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Lin H. Applying The Concept of Fuzzy Logic to Inverse Reinforcement Learning. [Thesis]. NSYSU; 2015. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-1025115-185021

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

University of Waterloo

23. Bhalla, Sushrut. Deep Multi Agent Reinforcement Learning for Autonomous Driving.

Degree: 2020, University of Waterloo

URL: http://hdl.handle.net/10012/15799

► Deep Learning and back-propagation have been successfully used to perform centralized training with communication protocols among multiple agents in a cooperative Multi-Agent Deep Reinforcement Learning… (more)

▼ Deep Learning and back-propagation have been successfully used to perform centralized training with communication protocols among multiple agents in a cooperative Multi-Agent Deep Reinforcement Learning (MARL) environment. In this work, I present techniques for centralized training of MARL agents in large scale environments and compare my work against current state of the art techniques. This work uses model-free Deep Q-Network (DQN) as the baseline model and allows inter agent communication for cooperative policy learning. I present two novel, scalable and centralized MARL training techniques (MA-MeSN, MA-BoN), which are developed under the principle that the behavior policy and message/communication policies have different optimization criteria. Thus, this work presents models which separate the message learning module from the behavior policy learning module. As shown in the experiments, the separation of these modules helps in faster convergence in complex domains like autonomous driving simulators and achieves better results than the current techniques in literature. Subsequently, this work presents two novel techniques for achieving decentralized execution for the communication based cooperative policy. The first technique uses behavior cloning as a method of cloning an expert cooperative policy to a decentralized agent without message sharing. In the second method, the behavior policy is coupled with a memory module which is local to each model. This memory model is used by the independent agents to mimic the communication policies of other agents and thus generate an independent behavior policy. This decentralized approach has minimal effect on degradation of the overall cumulative reward achieved by the centralized policy. Using a fully decentralized approach allows us to address the challenges of noise and communication bottlenecks in real-time communication channels. In this work, I theoretically and empirically compare the centralized and decentralized training algorithms to current research in the field of MARL. As part of this thesis, I also developed a large scale multi-agent testing environment. It is a new OpenAI-Gym environment which can be used for large scale multi-agent research as it simulates multiple autonomous cars driving cooperatively on a highway in the presence of a bad actor. I compare the performance of the centralized algorithms to existing state-of-the-art algorithms, for ex, DIAL and IMS which are based on cumulative reward achieved per episode and other metrics. MA-MeSN and MA-BoN achieve a cumulative reward of at-least 263% higher than the reward achieved by the DIAL and IMS. I also present an ablation study of the scalability of MA-BoN and show that MA-MeSN and MA-BoN algorithms only exhibit a linear increase in inference time and number of trainable parameters compared to quadratic increase for DIAL.

Subjects/Keywords: Machine Learning; Reinforcement Learning; Multi-Agent Reinforcement Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Bhalla, S. (2020). Deep Multi Agent Reinforcement Learning for Autonomous Driving. (Thesis). University of Waterloo. Retrieved from http://hdl.handle.net/10012/15799

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Bhalla, Sushrut. “Deep Multi Agent Reinforcement Learning for Autonomous Driving.” 2020. Thesis, University of Waterloo. Accessed March 04, 2021. http://hdl.handle.net/10012/15799.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Bhalla, Sushrut. “Deep Multi Agent Reinforcement Learning for Autonomous Driving.” 2020. Web. 04 Mar 2021.

Vancouver:

Bhalla S. Deep Multi Agent Reinforcement Learning for Autonomous Driving. [Internet] [Thesis]. University of Waterloo; 2020. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10012/15799.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Bhalla S. Deep Multi Agent Reinforcement Learning for Autonomous Driving. [Thesis]. University of Waterloo; 2020. Available from: http://hdl.handle.net/10012/15799

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Delft University of Technology

24. van der Wijden, R. (author). Preference-driven demonstrations ranking for inverse reinforcement learning.

Degree: 2016, Delft University of Technology

URL: http://resolver.tudelft.nl/uuid:4a85d32d-79da-4983-97d7-530c7bb1da98

►

New flexible teaching methods for robotics are needed to automate repetitive tasks that are currently still done by humans. For limited batch sizes, it is… (more)

▼

New flexible teaching methods for robotics are needed to automate repetitive tasks that are currently still done by humans. For limited batch sizes, it is too expensive to teach a robot a new task (Smith & Anderson, 2014). Ideally, such flexible robots can be taught a new task by a non-expert. A non-expert is a person who knows the task the robot should perform, but does not have experience in programming a robot. A powerful method that would allow for flexible robotics without the use of an expert is inverse reinforcement learning (IRL). IRL aims to learn the cost function out of demonstrations, this cost function is subsequently used to learn a policy which realizes the desired task. Current implementations focus more on the IRL algorithm itself and assume that there are enough demonstrations available and the quality of these demonstrations is also close enough to the optimal behaviour (Doerr et al., 2015). Whilst actually these demonstrations are very expensive and non-optimal. This thesis focuses on the effect of the quality of input demonstrations on the performance of the learned trajectory. Furthermore, how imperfect demonstrations still can be used, without lowering the performance of the learned trajectory. The first hypothesis is that the performance of the resulting trajectory depends on the average performance of the input demonstrations and the quantity of the input demonstrations has less of an effect. The second hypothesis is that by adding a ranking to the demonstrations, created through the preferences of non-robotic experts, the performance of the learned trajectory would be better than the average performance of the input demonstrations. The preferences of the non-robotic expert are collected through a crowdsourcing experiment. The preferences of the non-robotic expert are used to create an overall performance measurement. This overall performance measurement is used to obtain the sequentially order of the input demonstrations but also to evaluate the final learned trajectories. The results validate the first hypothesis. The average performance of the input demonstrations is determining the performance of the learned trajectory. The second hypothesis could not be confirmed. The results did not show any improvements in the performance of the learned trajectory when the ranking based on the preference of a non-robotic expert is added. It could be argued that the input demonstrations were too similar or the cost features used in IRL are not specific enough to create different cost functions and therefore create differently performing trajectories.

Mechanical, Maritime and Materials Engineering

Biomechanical Engineering

Advisors/Committee Members: Kober, J. (mentor).

Subjects/Keywords: robotics; reinforcement learning; preference learning; inverse reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

van der Wijden, R. (. (2016). Preference-driven demonstrations ranking for inverse reinforcement learning. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:4a85d32d-79da-4983-97d7-530c7bb1da98

Chicago Manual of Style (16^th Edition):

van der Wijden, R (author). “Preference-driven demonstrations ranking for inverse reinforcement learning.” 2016. Masters Thesis, Delft University of Technology. Accessed March 04, 2021. http://resolver.tudelft.nl/uuid:4a85d32d-79da-4983-97d7-530c7bb1da98.

MLA Handbook (7^th Edition):

van der Wijden, R (author). “Preference-driven demonstrations ranking for inverse reinforcement learning.” 2016. Web. 04 Mar 2021.

Vancouver:

van der Wijden R(. Preference-driven demonstrations ranking for inverse reinforcement learning. [Internet] [Masters thesis]. Delft University of Technology; 2016. [cited 2021 Mar 04]. Available from: http://resolver.tudelft.nl/uuid:4a85d32d-79da-4983-97d7-530c7bb1da98.

Council of Science Editors:

van der Wijden R(. Preference-driven demonstrations ranking for inverse reinforcement learning. [Masters Thesis]. Delft University of Technology; 2016. Available from: http://resolver.tudelft.nl/uuid:4a85d32d-79da-4983-97d7-530c7bb1da98

Oregon State University

25. Zhang, Wei, 1960-. Reinforcement learning for job-shop scheduling.

Degree: PhD, Computer Science, 1996, Oregon State University

URL: http://hdl.handle.net/1957/11721

Subjects/Keywords: Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Zhang, Wei, 1. (1996). Reinforcement learning for job-shop scheduling. (Doctoral Dissertation). Oregon State University. Retrieved from http://hdl.handle.net/1957/11721

Chicago Manual of Style (16^th Edition):

Zhang, Wei, 1960-. “Reinforcement learning for job-shop scheduling.” 1996. Doctoral Dissertation, Oregon State University. Accessed March 04, 2021. http://hdl.handle.net/1957/11721.

MLA Handbook (7^th Edition):

Zhang, Wei, 1960-. “Reinforcement learning for job-shop scheduling.” 1996. Web. 04 Mar 2021.

Vancouver:

Zhang, Wei 1. Reinforcement learning for job-shop scheduling. [Internet] [Doctoral dissertation]. Oregon State University; 1996. [cited 2021 Mar 04]. Available from: http://hdl.handle.net/1957/11721.

Council of Science Editors:

Zhang, Wei 1. Reinforcement learning for job-shop scheduling. [Doctoral Dissertation]. Oregon State University; 1996. Available from: http://hdl.handle.net/1957/11721

26. Clark, Kendrick Cheng Go. A Reinforcement Learning Model of the Shepherding Task : 羊飼い課題の強化学習モデル; ヒツジカイカダイノキョウカガクシュウモデル.

Degree: Nara Institute of Science and Technology / 奈良先端科学技術大学院大学

URL: http://hdl.handle.net/10061/10997

Subjects/Keywords: reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Clark, K. C. G. (n.d.). A Reinforcement Learning Model of the Shepherding Task : 羊飼い課題の強化学習モデル; ヒツジカイカダイノキョウカガクシュウモデル. (Thesis). Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Retrieved from http://hdl.handle.net/10061/10997

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Clark, Kendrick Cheng Go. “A Reinforcement Learning Model of the Shepherding Task : 羊飼い課題の強化学習モデル; ヒツジカイカダイノキョウカガクシュウモデル.” Thesis, Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Accessed March 04, 2021. http://hdl.handle.net/10061/10997.

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Clark, Kendrick Cheng Go. “A Reinforcement Learning Model of the Shepherding Task : 羊飼い課題の強化学習モデル; ヒツジカイカダイノキョウカガクシュウモデル.” Web. 04 Mar 2021.

Note: this citation may be lacking information needed for this citation format:
No year of publication.

Vancouver:

Clark KCG. A Reinforcement Learning Model of the Shepherding Task : 羊飼い課題の強化学習モデル; ヒツジカイカダイノキョウカガクシュウモデル. [Internet] [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10061/10997.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

Council of Science Editors:

Clark KCG. A Reinforcement Learning Model of the Shepherding Task : 羊飼い課題の強化学習モデル; ヒツジカイカダイノキョウカガクシュウモデル. [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; Available from: http://hdl.handle.net/10061/10997

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

27. Mauricio Alexandre Parente Burdelis. Temporal Difference Approach in Linearly Solvable Markov Decision Processes : 線形可解マルコフ決定過程における受動的ダイナミクスのモデリングと推定; センケイカカイマルコフケッテイカテイニオケルジュドウテキダイナミクスノモデリングトスイテイ.

Degree: 博士(工学), Nara Institute of Science and Technology / 奈良先端科学技術大学院大学

URL: http://hdl.handle.net/10061/9189

Subjects/Keywords: Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Burdelis, M. A. P. (n.d.). Temporal Difference Approach in Linearly Solvable Markov Decision Processes : 線形可解マルコフ決定過程における受動的ダイナミクスのモデリングと推定; センケイカカイマルコフケッテイカテイニオケルジュドウテキダイナミクスノモデリングトスイテイ. (Thesis). Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Retrieved from http://hdl.handle.net/10061/9189

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Burdelis, Mauricio Alexandre Parente. “Temporal Difference Approach in Linearly Solvable Markov Decision Processes : 線形可解マルコフ決定過程における受動的ダイナミクスのモデリングと推定; センケイカカイマルコフケッテイカテイニオケルジュドウテキダイナミクスノモデリングトスイテイ.” Thesis, Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Accessed March 04, 2021. http://hdl.handle.net/10061/9189.

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Burdelis, Mauricio Alexandre Parente. “Temporal Difference Approach in Linearly Solvable Markov Decision Processes : 線形可解マルコフ決定過程における受動的ダイナミクスのモデリングと推定; センケイカカイマルコフケッテイカテイニオケルジュドウテキダイナミクスノモデリングトスイテイ.” Web. 04 Mar 2021.

Note: this citation may be lacking information needed for this citation format:
No year of publication.

Vancouver:

Burdelis MAP. Temporal Difference Approach in Linearly Solvable Markov Decision Processes : 線形可解マルコフ決定過程における受動的ダイナミクスのモデリングと推定; センケイカカイマルコフケッテイカテイニオケルジュドウテキダイナミクスノモデリングトスイテイ. [Internet] [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10061/9189.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

Council of Science Editors:

Burdelis MAP. Temporal Difference Approach in Linearly Solvable Markov Decision Processes : 線形可解マルコフ決定過程における受動的ダイナミクスのモデリングと推定; センケイカカイマルコフケッテイカテイニオケルジュドウテキダイナミクスノモデリングトスイテイ. [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; Available from: http://hdl.handle.net/10061/9189

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

28. 森本, 淳. Hierarchical Decomposition and Min-max Strategy for Fast and Robust Reinforcement Learning in the Real Environment : 階層分割とMin-max戦略による実環境での高速かつロバストな強化学習; カイソウブンカツト Min-max センリャクニヨルジツカンキョウデノコウソクカツロバストナキョウカガクシュウ.

Degree: Nara Institute of Science and Technology / 奈良先端科学技術大学院大学

URL: http://hdl.handle.net/10061/2966

Subjects/Keywords: reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

森本, �. (n.d.). Hierarchical Decomposition and Min-max Strategy for Fast and Robust Reinforcement Learning in the Real Environment : 階層分割とMin-max戦略による実環境での高速かつロバストな強化学習; カイソウブンカツト Min-max センリャクニヨルジツカンキョウデノコウソクカツロバストナキョウカガクシュウ. (Thesis). Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Retrieved from http://hdl.handle.net/10061/2966

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

森本, 淳. “Hierarchical Decomposition and Min-max Strategy for Fast and Robust Reinforcement Learning in the Real Environment : 階層分割とMin-max戦略による実環境での高速かつロバストな強化学習; カイソウブンカツト Min-max センリャクニヨルジツカンキョウデノコウソクカツロバストナキョウカガクシュウ.” Thesis, Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Accessed March 04, 2021. http://hdl.handle.net/10061/2966.

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

森本, 淳. “Hierarchical Decomposition and Min-max Strategy for Fast and Robust Reinforcement Learning in the Real Environment : 階層分割とMin-max戦略による実環境での高速かつロバストな強化学習; カイソウブンカツト Min-max センリャクニヨルジツカンキョウデノコウソクカツロバストナキョウカガクシュウ.” Web. 04 Mar 2021.

Note: this citation may be lacking information needed for this citation format:
No year of publication.

Vancouver:

森本 �. Hierarchical Decomposition and Min-max Strategy for Fast and Robust Reinforcement Learning in the Real Environment : 階層分割とMin-max戦略による実環境での高速かつロバストな強化学習; カイソウブンカツト Min-max センリャクニヨルジツカンキョウデノコウソクカツロバストナキョウカガクシュウ. [Internet] [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10061/2966.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

Council of Science Editors:

森本 �. Hierarchical Decomposition and Min-max Strategy for Fast and Robust Reinforcement Learning in the Real Environment : 階層分割とMin-max戦略による実環境での高速かつロバストな強化学習; カイソウブンカツト Min-max センリャクニヨルジツカンキョウデノコウソクカツロバストナキョウカガクシュウ. [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; Available from: http://hdl.handle.net/10061/2966

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

29. 南條, 信人. 動的な部分空間生成による価値の逐次推定を行う強化学習法 : An effective reinforcement learning with automatic construction of basis functions and sequential approximation; ドウテキナブブンクウカンセイセイニヨルカチノチクジスイテイオオコナウキョウカガクシュウホウ.

Degree: Nara Institute of Science and Technology / 奈良先端科学技術大学院大学

URL: http://hdl.handle.net/10061/4584

Subjects/Keywords: Reinforcement learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

南條, �. (n.d.). 動的な部分空間生成による価値の逐次推定を行う強化学習法 : An effective reinforcement learning with automatic construction of basis functions and sequential approximation; ドウテキナブブンクウカンセイセイニヨルカチノチクジスイテイオオコナウキョウカガクシュウホウ. (Thesis). Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Retrieved from http://hdl.handle.net/10061/4584

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

南條, 信人. “動的な部分空間生成による価値の逐次推定を行う強化学習法 : An effective reinforcement learning with automatic construction of basis functions and sequential approximation; ドウテキナブブンクウカンセイセイニヨルカチノチクジスイテイオオコナウキョウカガクシュウホウ.” Thesis, Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Accessed March 04, 2021. http://hdl.handle.net/10061/4584.

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

南條, 信人. “動的な部分空間生成による価値の逐次推定を行う強化学習法 : An effective reinforcement learning with automatic construction of basis functions and sequential approximation; ドウテキナブブンクウカンセイセイニヨルカチノチクジスイテイオオコナウキョウカガクシュウホウ.” Web. 04 Mar 2021.

Note: this citation may be lacking information needed for this citation format:
No year of publication.

Vancouver:

南條 �. 動的な部分空間生成による価値の逐次推定を行う強化学習法 : An effective reinforcement learning with automatic construction of basis functions and sequential approximation; ドウテキナブブンクウカンセイセイニヨルカチノチクジスイテイオオコナウキョウカガクシュウホウ. [Internet] [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10061/4584.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

Council of Science Editors:

南條 �. 動的な部分空間生成による価値の逐次推定を行う強化学習法 : An effective reinforcement learning with automatic construction of basis functions and sequential approximation; ドウテキナブブンクウカンセイセイニヨルカチノチクジスイテイオオコナウキョウカガクシュウホウ. [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; Available from: http://hdl.handle.net/10061/4584

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

30. Rodrigues, Alan de Souza. Model-Free and Model-Based Reinforcement Learning Strategies in the Acquisition of Sequential Behaviors : 系列運動の獲得におけるモデルフリーとモデルベース強化学習戦略; ケイレツウンドウノカクトクニオケルモデルフリートモデルベースキョウカガクシュウセンリャク.

Degree: Nara Institute of Science and Technology / 奈良先端科学技術大学院大学

URL: http://hdl.handle.net/10061/4681

Subjects/Keywords: Reinforcement Learning

Record Details Similar Records Cite « Share

❌

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

APA (6^th Edition):

Rodrigues, A. d. S. (n.d.). Model-Free and Model-Based Reinforcement Learning Strategies in the Acquisition of Sequential Behaviors : 系列運動の獲得におけるモデルフリーとモデルベース強化学習戦略; ケイレツウンドウノカクトクニオケルモデルフリートモデルベースキョウカガクシュウセンリャク. (Thesis). Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Retrieved from http://hdl.handle.net/10061/4681

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16^th Edition):

Rodrigues, Alan de Souza. “Model-Free and Model-Based Reinforcement Learning Strategies in the Acquisition of Sequential Behaviors : 系列運動の獲得におけるモデルフリーとモデルベース強化学習戦略; ケイレツウンドウノカクトクニオケルモデルフリートモデルベースキョウカガクシュウセンリャク.” Thesis, Nara Institute of Science and Technology / 奈良先端科学技術大学院大学. Accessed March 04, 2021. http://hdl.handle.net/10061/4681.

Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7^th Edition):

Rodrigues, Alan de Souza. “Model-Free and Model-Based Reinforcement Learning Strategies in the Acquisition of Sequential Behaviors : 系列運動の獲得におけるモデルフリーとモデルベース強化学習戦略; ケイレツウンドウノカクトクニオケルモデルフリートモデルベースキョウカガクシュウセンリャク.” Web. 04 Mar 2021.

Note: this citation may be lacking information needed for this citation format:
No year of publication.

Vancouver:

Rodrigues AdS. Model-Free and Model-Based Reinforcement Learning Strategies in the Acquisition of Sequential Behaviors : 系列運動の獲得におけるモデルフリーとモデルベース強化学習戦略; ケイレツウンドウノカクトクニオケルモデルフリートモデルベースキョウカガクシュウセンリャク. [Internet] [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; [cited 2021 Mar 04]. Available from: http://hdl.handle.net/10061/4681.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

Council of Science Editors:

Rodrigues AdS. Model-Free and Model-Based Reinforcement Learning Strategies in the Acquisition of Sequential Behaviors : 系列運動の獲得におけるモデルフリーとモデルベース強化学習戦略; ケイレツウンドウノカクトクニオケルモデルフリートモデルベースキョウカガクシュウセンリャク. [Thesis]. Nara Institute of Science and Technology / 奈良先端科学技術大学院大学; Available from: http://hdl.handle.net/10061/4681

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.

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

Open Access Theses and Dissertations