Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 58 for “"Reward Shaping"”.

  1. REWARD SHAPING FOR REINFORCEMENT LEARNING

    … Learning often suffers from sparse and delayed rewards in real-world environments, which severely hinders learning efficiency. This dissertation systematically studies reward shaping (RS) frameworks to improve exploration, stability, and sample efficiency under limited native reward signals. We …

    nus Repository record for REWARD SHAPING FOR REINFORCEMENT LEARNING (opens in a new tab) Download file for REWARD SHAPING FOR REINFORCEMENT LEARNING (opens in a new tab)

  2. Automating abstraction for potential-based reward shaping

    … have been observed throughout the literature. Reward Shaping is one such technique for utilising abstractions in this way. This thesis focuses on how an agent can learn its own abstractions from its own experiences to be used for Potential Based Reward Shaping. As the thesis progresses, the …

    whiterose Download PDF for Automating abstraction for potential-based reward shaping (opens in a new tab)

  3. Theory and Application of Reward Shaping in Reinforcement Learning

    … predictions; most notably that reducing the reward horizon implies faster learning. The bipedal walking task demonstrates that our reward shaping techniques allow a conventional reinforcement learning algorithm to find a good behavior efficiently despite a large state space with stochastic …

    uiuc Repository record for Theory and Application of Reward Shaping in Reinforcement Learning (opens in a new tab)

  4. ROBUST AND AGILE QUADRUPED ​CONTROL VIA REWARD-SHAPING REINFORCEMENT LEARNING

    … (RL) has emerged as a prominent approach, and reward-shaping is essential in robot training. To address this, our work introduces two unified reward design frameworks, each tailored to distinct task categories. In the first part, we present a teacher–student RL architecture for fault-tolerant …

    nus Repository record for ROBUST AND AGILE QUADRUPED ​CONTROL VIA REWARD-SHAPING REINFORCEMENT LEARNING (opens in a new tab) Download file for ROBUST AND AGILE QUADRUPED ​CONTROL VIA REWARD-SHAPING REINFORCEMENT LEARNING (opens in a new tab)

  5. Knowledge-Based Reward Shaping with Knowledge Revision in Reinforcement Learning

    … feedback about its behaviour in terms of rewards through constant interaction with the environment and in time manages to identify which actions are more beneficial for each situation. Typically reinforcement learning assumes the agent has no prior knowledge about the environment it is …

    whiterose Download PDF for Knowledge-Based Reward Shaping with Knowledge Revision in Reinforcement Learning (opens in a new tab)

  6. Reward Shaping in Single and Multi-Agent Deep Reinforcement Learning

    … effectively complete tasks, especially when the reward provided by the environment is sparse or significantly delayed. In these situations, the agent will not be able to receive immediate feedback on its actions. This is called the credit assignment problem. In many scenarios, it is hard to …

    washington Repository record for Reward Shaping in Single and Multi-Agent Deep Reinforcement Learning (opens in a new tab)

  7. Advancing reasoning and planning in large language models via reward shaping

    … reasoning and planning abilities of LLMs through reward shaping to guide LLM decision-making by optimizing rewards for desired outcomes. The core contributions of this thesis are organized around three key aspects of effective and robust reasoning in LLM agents: (1) advancing LLM reasoning and …

    gatech Repository record for Advancing reasoning and planning in large language models via reward shaping (opens in a new tab)

  8. Potential-Based Reward Shaping for Knowledge-Based, Multi-Agent Reinforcement Learning

    … suitable state representation and an informative reward function is guaranteed to learn how to maximise its long term reward. Incorporating domain knowledge, typically known by the system designer, can minimise the number of suboptimal behaviours tried and, therefore, speed up the rate of …

    whiterose Download PDF for Potential-Based Reward Shaping for Knowledge-Based, Multi-Agent Reinforcement Learning (opens in a new tab)

  9. Optimizing Reinforcement Learning: Fog and Edge Resource Management Through Bootstrapping and Reward Shaping

    … Network solution as a potential-based reward shaping solution; (5) developing another reward shaping solution based on Convolutional Neural Network (CNN) carefully designed and inspired by the value iteration network (VIN), to speed learning. Besides these contributions, we present a …

    concordia Download PDF for Optimizing Reinforcement Learning: Fog and Edge Resource Management Through Bootstrapping and Reward Shaping (opens in a new tab)

  10. Learning Expressive Quadrupedal Locomotion Guided by Kinematic Trajectory Generator

    … and a reference-based approach using solely reward shaping, i.e. specification of the motion through the reward, have been implemented. They produced walking gaits in simulation. Yet, the motions produced by the reference-based approach had limited footstep height and balance issues. The …

    gatech Repository record for Learning Expressive Quadrupedal Locomotion Guided by Kinematic Trajectory Generator (opens in a new tab)

  11. Knowledge and Ignorance in Reinforcement Learning

    … three: the state-space/transition model, the reward function, and the observation model. In this thesis, I present a framework for studying how the state of knowledge or uncertainty of each component affects the Reinforcement Learning process. I focus on the reward function and the observation …

    uiuc Repository record for Knowledge and Ignorance in Reinforcement Learning (opens in a new tab)

  12. Improving Exploration in Reinforcement Learning through Domain Knowledge and Parameter Analysis

    … actions in order to maximise the long term reward. Solving this type of problems is a hard task when real domains of realistic size are considered because the state space grows exponentially with each state feature added to the representation of the problem. In its basic form, reinforcement …

    whiterose Download PDF for Improving Exploration in Reinforcement Learning through Domain Knowledge and Parameter Analysis (opens in a new tab)

  13. Deep learning techniques for financial data

    … to optimize their trading performance. The reward signal in these approaches is the achieved trading profit. Novel Deep Reinforcement Learning (DRL) methods are proposed for Reward Shaping and Knowledge Distillation. In the proposed Reward Shaping method, we implement an auxiliary reward, …

    greece Repository record for Deep learning techniques for financial data (opens in a new tab)

  14. Rational communication for the coordination of multi-Agent systems

    … from the field of Machine Learning, namely Reward Shaping, which allows the decentralised POMDP to be transformed into individual agent POMDPs that can be solved more easily. This approach can use a heuristic transformation to allow the approach to work in large problems like RobocupRescue …

    soton Download PDF for Rational communication for the coordination of multi-Agent systems (opens in a new tab)

  15. LEARNING GENERALIZABLE REPRESENTATIONS IN REINFORCEMENT LEARNING

    … the agent more quickly adapt to new tasks in reward shaping. Third, we introduce a mixture regularization to shape the learned representation. On the Procgen benchmark, the mixture regularization greatly boosts the agent’s zero-shot transfer performance. We believe our findings and methods …

    nus Repository record for LEARNING GENERALIZABLE REPRESENTATIONS IN REINFORCEMENT LEARNING (opens in a new tab) Download file for LEARNING GENERALIZABLE REPRESENTATIONS IN REINFORCEMENT LEARNING (opens in a new tab)

  16. ATTENTIVE RECURRENT NEURAL NETWORKS

    … be derived by either reinforcement learning with reward shaping or direct probabilistic inference via importance sampling. Finally, Partially Observable Trust Region Policy Optimization (PO-TRPO) algorithm is developed to learn navigation policies for socially concomitant mobile robot navigation. …

    nus Repository record for ATTENTIVE RECURRENT NEURAL NETWORKS (opens in a new tab) Download file for ATTENTIVE RECURRENT NEURAL NETWORKS (opens in a new tab)

  17. Proximal Policy Optimization for Formation Control and Obstacle Avoidance in Multi-Agent Systems

    … avoidance in multi-agent systems through reward shaping. The objective of this work is to study the application of proximal policy optimization (PPO) algorithm for maneuvering a formation of agents around obstacles. Each agent in the multi-agent system is modeled as a holonomic …

    concordia Download PDF for Proximal Policy Optimization for Formation Control and Obstacle Avoidance in Multi-Agent Systems (opens in a new tab)

  18. Learning-Based Task Assignment for Automated Guided Vehicles: Applying graph neural networks to optimise task assignment in an online warehouse environment

    … of deployment, specifically the complexities of reward shaping in real-world logistics systems and the systemic barriers to integrating learning-based methods into legacy industrial infrastructures.

    chalmers Repository record for Learning-Based Task Assignment for Automated Guided Vehicles: Applying graph neural networks to optimise task assignment in an online warehouse environment (opens in a new tab)

  19. Evoluční algoritmy v posilovaném učení

    … k neželaným javom, ako je zneužívanie odmeny (reward hacking) alebo úplné zlyhanie učenia. Formovanie odmien (reward shaping) je metóda, ktorá agentovi poskytuje dodatočné informácie o úlohe s cieľom zlepšiť efektivitu a stabilitu jeho učenia. Táto diplomová práca sa zaoberá použitím …

    brno-tech Repository record for Evoluční algoritmy v posilovaném učení (opens in a new tab)

  20. ASH: Agents that self-hone via Embodied Learning

    … AI, as current methods rely on hand-engineered rewards or action-labeled demonstrations, neither of which scales. We introduce ASH, an agentic system that learns an embodied policy from unlabeled, noisy internet video, without reward shaping or expert annotation. ASH follows a self improvement …

    waterloo Repository record for ASH: Agents that self-hone via Embodied Learning (opens in a new tab)

Page 1 of 3