Safe exploration for reinforcement learning

13 years 5 months ago

Download ahans.de

In this paper we define and address the problem of safe exploration in the context of reinforcement learning. Our notion of safety is concerned with states or transitions that can lead to damage and thus must be avoided. We introduce the concepts of a safety function for determining a state's safety degree and that of a backup policy that is able to lead the system under control from a critical state back to a safe one. Moreover, we present a level-based exploration scheme that is able to generate a comprehensive base of observations while adhering safety constraints.We evaluate our approach on a simplified simulation of a gas turbine.

Alexander Hans, Daniel Schneegaß, Anton Maxi

Real-time Traffic

ESANN 2008 | Level-based Exploration Scheme | Neural Networks | Safety | State's Safety Degree |

claim paper

» Lyapunov Design for Safe Reinforcement Learning

» Generalization in Reinforcement Learning Safely Approximating the Value Function

» Safe State Abstraction and Reusable Continuing Subtasks in Hierarchical Reinforcement Lear...

» Modelbased reinforcement learning with nearly tight exploration complexity bounds

» Coordinated exploration in multiagent reinforcement learning an application to loadbalanci...

» Action Inhibition

» Efficient exploration through active learning for value function approximation in reinforc...

» Exploration and apprenticeship learning in reinforcement learning

Post Info
More Details (n/a)

Added	29 Oct 2010
Updated	29 Oct 2010
Type	Conference
Year	2008
Where	ESANN
Authors	Alexander Hans, Daniel Schneegaß, Anton Maximilian Schäfer, Steffen Udluft

Comments (0)

Sciweavers

Safe exploration for reinforcement learning

ESANN 2008 | Level-based Exploration Scheme | Neural Networks | Safety | State's Safety Degree |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers