Almost surely safe exploration and exploitation for deep reinforcement learning with state safety estimation
This study aims to address the challenge of constrained reinforcement learning (CRL),
which seeks to maximize cumulative rewards while making agents avoid risks. Existing CRL …
which seeks to maximize cumulative rewards while making agents avoid risks. Existing CRL …
Planning under uncertainty for safe robot exploration using Gaussian process prediction
The exploration of new environments is a crucial challenge for mobile robots. This task
becomes even more complex with the added requirement of ensuring safety. Here, safety …
becomes even more complex with the added requirement of ensuring safety. Here, safety …
Transductive active learning: Theory and applications
We study a generalization of classical active learning to real-world settings with concrete
prediction targets where sampling is restricted to an accessible region of the domain, while …
prediction targets where sampling is restricted to an accessible region of the domain, while …
Lipschitz safe Bayesian optimization for automotive control
J Menn, P Pelizzari, M Fleps-Dezasse… - arXiv preprint arXiv …, 2025 - arxiv.org
Controller tuning is a labor-intensive process that requires human intervention and expert
knowledge. Bayesian optimization has been applied successfully in different fields to …
knowledge. Bayesian optimization has been applied successfully in different fields to …
Pluck and Play: Self-supervised Exploration of Chordophones for Robotic Playing
M Görner, N Hendrich, J Zhang - 2024 IEEE International …, 2024 - ieeexplore.ieee.org
Existing robotic musicians utilize detailed handcrafted instrument models to generate or
learn policies for playing because model-free or inaccurate policy rollouts might easily …
learn policies for playing because model-free or inaccurate policy rollouts might easily …
Transductive active learning with application to safe bayesian optimization
Safe Bayesian optimization (Safe BO) is the task of learning an optimal policy within an
unknown environment, while ensuring that safety constraints are not violated. We analyze …
unknown environment, while ensuring that safety constraints are not violated. We analyze …
Information-based Transductive Active Learning
We generalize active learning to address real-world settings where sampling is restricted to
an accessible region of the domain, while prediction targets may lie outside this region. To …
an accessible region of the domain, while prediction targets may lie outside this region. To …
Information-Theoretic Safe Bayesian Optimization
We consider a sequential decision making task, where the goal is to optimize an unknown
function without evaluating parameters that violate an a~ priori unknown (safety) constraint …
function without evaluating parameters that violate an a~ priori unknown (safety) constraint …
Gaussian Processes in Control: Performance Guarantees through Efficient Learning
A Lederer - 2023 - mediatum.ub.tum.de
While Gaussian process (GP) models promise to enable learning control, their application
creates theoretical and practical issues. In this thesis, theoretical challenges are addressed …
creates theoretical and practical issues. In this thesis, theoretical challenges are addressed …