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 182 for “"Sample Efficiency"”.

  1. Evaluation of sample efficiency in offline reinforcement learning

    … of the training set size and thus by their data efficiency. Sequential evaluation provides valuable insights into the data efficiency of the learning process and the robustness of algorithms to distribution changes in the dataset while also harmonizing the visualization of the offline and online …

    ets-quebec Download PDF for Evaluation of sample efficiency in offline reinforcement learning (opens in a new tab)

  2. ENHANCING SAMPLE EFFICIENCY AND PRIVACY IN REAL-WORLD DECISION-MAKING: A FEDERATED REINFORCEMENT LEARNING APPROACH

    … a suite of federated RL frameworks that improve sample efficiency and safeguard privacy. We introduce FedPG‐BR, a variance‐reduced federated policy gradient method endowed with Byzantine resilience; its theoretical guarantees ensure that sample efficiency improves with more participating agents …

    nus Repository record for ENHANCING SAMPLE EFFICIENCY AND PRIVACY IN REAL-WORLD DECISION-MAKING: A FEDERATED REINFORCEMENT LEARNING APPROACH (opens in a new tab) Download file for ENHANCING SAMPLE EFFICIENCY AND PRIVACY IN REAL-WORLD DECISION-MAKING: A FEDERATED REINFORCEMENT LEARNING APPROACH (opens in a new tab)

  3. Improving sample-efficiency of model-free reinforcement learning algorithms on image inputs with representation learning

    … representations and policies result in higher sample-efficiency if adequate regularization is provided. We further hypothesize that representations which correlate more strongly with the underlying Markov decision process result in additional sample-efficiency. These hypotheses are tested …

    chalmers Repository record for Improving sample-efficiency of model-free reinforcement learning algorithms on image inputs with representation learning (opens in a new tab)

  4. Practical reinforcement learning for adaptive robotic manipulation: Sample efficiency, sim-to-real transfer, and context inference

    … to adapt to novel tasks and environments in a sample-efficient and robust manner. A framework is proposed to enable such adaptability through three interrelated and complementary contributions in the field of reinforcement learning (RL) for robot manipulation. The central challenges …

    uvic Repository record for Practical reinforcement learning for adaptive robotic manipulation: Sample efficiency, sim-to-real transfer, and context inference (opens in a new tab)

  5. Advancing Efficiency and Safety in Autonomous Sequential Decision Making

    … substantial obstacles, chiefly in achieving sample (data) efficiency and ensuring agent safety in unpredictable, dynamic environments. Additionally, the inherent partial knowledge due to sensory and model limitations complicates agents' functionality in complex scenarios. This thesis aims to …

    toronto-retro Repository record for Advancing Efficiency and Safety in Autonomous Sequential Decision Making (opens in a new tab)

  6. Towards Sample-Efficient Deep Reinforcement Learning

    … require a prohibitively large number of data samples to train effective strategies. Consequently, it is difficult to apply these DRL algorithms to complex real-world problems that require high costs to collect a large volume of data samples. This dissertation proposes new mechanisms to address …

    vcu Repository record for Towards Sample-Efficient Deep Reinforcement Learning (opens in a new tab)

  7. Bridging control and reinforcement learning with partial model knowledge

    … is flexible yet typically suffers from poor sample efficiency. The proposed methods integrate the strengths of both paradigms by combining model-based control where knowledge is available with learning-based adaptation for unknown components. We first consider linear systems and introduce …

    ubc Repository record for Bridging control and reinforcement learning with partial model knowledge (opens in a new tab) Download PDF for Bridging control and reinforcement learning with partial model knowledge (opens in a new tab)

  8. Efficient Methods for Deep Reinforcement Learning: Algorithms and Applications

    … However, deep RL agents are notoriously sample-inefficient; they often need to collect a large number of samples from the environment to achieve a reasonable performance. This sample efficiency issue becomes more pronounced in sparse reward environments, where the rewards are zeros in …

    rice Repository record for Efficient Methods for Deep Reinforcement Learning: Algorithms and Applications (opens in a new tab)

  9. Evolutionary Policy Optimization

    … trade-off without sacrificing sample efficiency. Policy gradient (PG) methods excel in exploitation through fine-grained, gradient-based optimization but often struggle with exploration due to their focus on local search. In contrast, evolutionary computation (EC) methods excel …

    texas Repository record for Evolutionary Policy Optimization (opens in a new tab)

  10. Reward Machines

    … use of reinforcement learning in the real world: sample efficiency and partial observability. In this dissertation, we tackle these two problems through the creation of reward machines. Reward machines are automata-based representations of a reward function that expose reward structures to the …

    toronto-retro Repository record for Reward Machines (opens in a new tab)

  11. SAMPLE-EFFICIENT AUTOMATED MACHINE LEARNING WITH BAYESIAN OPTIMIZATION

    … which can be attributed to its impressive sample efficiency and theoretical convergence guarantee. Despite recent advances, there are still important scenarios where we can further improve the sample efficiency of BO for AutoML by exploiting naturally available auxiliary information, or …

    nus Repository record for SAMPLE-EFFICIENT AUTOMATED MACHINE LEARNING WITH BAYESIAN OPTIMIZATION (opens in a new tab) Download file for SAMPLE-EFFICIENT AUTOMATED MACHINE LEARNING WITH BAYESIAN OPTIMIZATION (opens in a new tab)

  12. Leveraging physics-based knowledge for efficient real-world reinforcement learning

    … that embed physical principles to enhance sample efficiency for real-world applications. Two key frameworks are presented: a policy optimization method incorporating simplified physics models and low-level feedback for stable learning, and a model-based approach that combines structured …

    texas Repository record for Leveraging physics-based knowledge for efficient real-world reinforcement learning (opens in a new tab)

  13. On the Value of Online Learning for Cognitive Radar Waveform Selection

    … inherent to the waveform selection problem, sample-efficiency and universality. Sample-efficiency corresponds to the number of experiences a learning algorithm requires to achieve desirable performance. Universality refers to the learning algorithm's ability to achieve desirable performance …

    vt Repository record for On the Value of Online Learning for Cognitive Radar Waveform Selection (opens in a new tab)

  14. Reinforcement learning in large state action spaces

    … These include difficulties with exploration, low sample efficiency, computational intractability, task constraints like decentralization and lack of guarantees about important properties like performance, generalization and robustness in potentially unseen scenarios.</p> <p>This thesis is …

    oxford Repository record for Reinforcement learning in large state action spaces (opens in a new tab)

  15. Improving single and multi-agent deep reinforcement learning methods

    … to better performance of agents, improved sample efficiency, and generalizability of learned agent policies.</p> <p>In Part I of the thesis, we focus on exploration in single-agent RL settings where an agent must interact with a complex environment to pursue a goal. An agent that fails to …

    oxford Repository record for Improving single and multi-agent deep reinforcement learning methods (opens in a new tab)

  16. AUTOMATED QUALITY EVALUATION OF CHINESE ONLINE HEALTH INFORMATION USING THE DISCERN SCALE: A DAL FRAMEWORK STUDY

    … label structure, annotation reliability, and sample efficiency. A systematic, governance oriented framework for automated OHI quality evaluation is therefore needed. Methods. This dissertation proposes a closed loop DAL framework (Dataset governance, Algorithm development, and Learning curve …

    jhu Repository record for AUTOMATED QUALITY EVALUATION OF CHINESE ONLINE HEALTH INFORMATION USING THE DISCERN SCALE: A DAL FRAMEWORK STUDY (opens in a new tab) Embargoed until 2028-05-01

  17. World modeling and black-box optimization with language models

    … thesis explores how language can improve the sample efficiency and generalization of RL agents by incorporating prior knowledge into their internal world models. It further investigates the use of large language models (LLMs) for black-box optimization (BBO), where an unknown objective must be …

    aalto Repository record for World modeling and black-box optimization with language models (opens in a new tab)

  18. Capability-Aware Shared Hypernetworks for Heterogeneous Multi-Agent Coordination

    … behaviors within a single architecture for sample-efficiency, but are limited in their ability to learn diverse behaviors; ii) independent methods, which learn a separate policy for each agent, enabling greater diversity at the cost of sample- and parameter-efficiency. Prior work on learning …

    gatech Repository record for Capability-Aware Shared Hypernetworks for Heterogeneous Multi-Agent Coordination (opens in a new tab)

  19. Online Learning and Planning of Dynamical Systems Using Gaussian Processes

    … with no task-specific knowledge, it often lacks efficiency in terms of the number of required samples. This lack of sample efficiency makes RL inapplicable to many real world problems. Thus, a central challenge in RL is how to extract more information from available experience to facilitate fast …

    nus Repository record for Online Learning and Planning of Dynamical Systems Using Gaussian Processes (opens in a new tab) Download file for Online Learning and Planning of Dynamical Systems Using Gaussian Processes (opens in a new tab)

  20. Transformer-Based Models for Identifying Customer Needs in User-Generated Content: Performance Gaps, Unintended Bias, and Broader Implications

    … for generalizability, robustness, fairness, and sample efficiency across varied settings. Results indicate competitive accuracy, with gains in F1 up to 18% over baselines, but also consistent limitations: shared error patterns, difficulty with rare or unseen needs, reliance on lexical cues that …

    concordia Download PDF for Transformer-Based Models for Identifying Customer Needs in User-Generated Content: Performance Gaps, Unintended Bias, and Broader Implications (opens in a new tab)

Page 1 of 10