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 38 for “"Safe reinforcement learning"”.
-
Safe reinforcement learning in automotive
This work applies Logically Constrained Reinforcement Learning (LCRL) frame-work for synthesizing policies for active monitoring and management of sensor modules in autonomous vehicles. LCRL allows synthesizing policies for unknown and continuous-state Markov Decision Processes (MDPs) such that a …
-
Safe Reinforcement Learning Using Formally Verified Abstract Policies
Reinforcement learning (RL) is an artificial intelligence technique for finding optimal solutions for sequential decision-making problems modelled as Markov decision processes (MDPs). Objectives are represented as numerical rewards in the model where positive values represent achievements and …
-
The Development of Robust and Safe Reinforcement Learning Methods
Reinforcement learning (RL) has become a prominent research area in recent years due to its ability to address decision-making problems by maximizing cumulative rewards in complex environments through interactions with the environment. Concurrently, the rapid progress in deep neural network has …
-
SAFE REINFORCEMENT LEARNING-BASED GREEN LIGHT OPTIMAL SPEED ADVISORY FOR MIXED-TRAFFIC PLATOONS
This thesis develops a platoon-centric, safe RL-based Green Light Optimal Speed Advisory (GLOSA) system to optimize the CAV speed profile of a mixed-traffic platoon. First, we design a multi-agent RL algorithm to achieve a balance between the energy and travel efficiency of a mixed-traffic platoon, …
-
Safe reinforcement learning: An overview, a hybrid systems perspective, and a case study
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 novel policies that may not be easily expressed with rules-based strategies or traditional control techniques. Over the years since …
-
A Safe Reinforcement Learning Approach for Energy Efficient Federated Learning in Wireless Communication Networks
Με την συνεχώς αυξανόμενη πολυπλοκότητα των επερχόμενων κινητών δικτύων, λόγω της εξάπλωσης των ετερογενών συνδεδεμένων συσκευών και της συνεχούς αύξησης της υπολογιστικής απόδοσης, η ύπαρξη ενός πλαισίου Τεχνητής Νοημοσύνης (ΤΝ) για την αυτοματοποίηση του δικτυακού συστήματος είναι μείζονος …
-
Safe and transferable multi-task bandit learning with shared representations and safe reinforcement learning
… Over the past decade, federated and multi-task learning have emerged as powerful collaborative paradigms, significantly improving learning efficiency by facilitating joint learning in a privacy-preserved manner. The most promising direction for learning-based control is a dynamic sequential …
-
Meta-optimization in safe reinforcement learning: Enhancing safety at training and deployment with fewer hyperparameters
Reinforcement learning (RL) is a trial-and-error framework for enabling intelligent systems to learn the optimal behaviour based on the feedback from the environment. In recent years, successful application of RL in controlling various embodied systems have been observed. However, the real-world …
-
Uncertainty Estimation Toward Safe Ai
Safety critical AI systems interact with environments based on inductively learned predictors, which may not be always correct. To complement the incorrect predictions, quantifying uncertainty on predictions is crucial to guarantee the safety of the AI systems. The major challenge of uncertainty …
-
Machine Learning for Agile Robotic Control
… tempting to fully forego structure in favor of learning-based methods for robotics, we show how data and learning can be gracefully incorporated in a structured way. In particular, we focus on the control setting, and we demonstrate that robotic control offers a variety of modes that data can be …
-
Closed-loop Real-time Safety Critic as Guardrail in Robotics
… and manipulation settings, where physical safety is a basic premise rather than an optional supplement. While robot policies may be obtained through reinforcement learning, imitation learning, optimal control, etc., the development normally involves training, simulation, testing, or rollout …
-
Trustworthy Autonomy Through Robust Control and Alignment
… intelligence systems are increasingly applied in safety critical domains such as robotics, autonomous driving, and decision making under uncertainty, ensuring their trustworthiness has become a central challenge. This dissertation addresses two major facets of trustworthy AI: reliable control …
-
Online data-driven control of safety-critical systems
… systems have brought questions related to safety and adaptation to the forefront of the minds of controls and robotics engineers. Often, such autonomous systems are deemed to be safety-critical in the sense that failures during operation could significantly harm the system itself, other …
-
Task-aware planning and learning in partially observable environments
… there is a requirement to ensure that they are safe. We present theory and algorithms for decision-making agents making verifiably correct decisions under incomplete information. In particular, we provide a method that integrates scalable recurrent neural network representations from machine …
-
Electric Bus Charging Scheduling based on Deep Reinforcement Learning
… this thesis develops novel Hierarchical Deep Reinforcement Learning (HDRL) frameworks combined with Safe Reinforcement Learning (Safe RL) methods. The primary objective is to devise intelligent, scalable, and safety-aware charging scheduling algorithms tailored for practical, uncertain …
-
Leveraging Constraint Space Manifolds and Lottery Tickets to Learn Complex Control Tasks with Pruned Neural Networks
… problems in Robotics. The application of Deep Reinforcement Learning (Deep-RL) to these problems has enabled far more dynamic and adaptive solutions than traditional model-based approaches. In recent years, considerable research attention has shifted towards safe reinforcement learning, where …
-
Neurosymbolic approaches to safe machine learning
… substantial risk, and in order to deploy machine learning systems in the real world, we need tools to analyze the behavior of these systems. This presents a problem to researchers because neural networks are generally resistant to traditional approaches to program analysis. From a formal analysis …
Page 1 of 2