MétaCan
← all works

Safe Model-based Reinforcement Learning with Stability Guarantees

2017· article· en· 337 citations· W2618318883 on OpenAlex· 10.48550/arxiv.1705.08551

Why is this work in the frame?

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

Canadian funderA Canadian agency funded it. The work may carry no Canadian affiliation at all.

No Canadian affiliation. An affiliation-only frame — the usual design — would never have seen this work. It is one of the works that make the case for inverting the frame.

Machine scores (provisional)

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Opus teacher head0.061
GPT teacher head0.200
Teacher spread
0.139 · how far apart the two teachers sit on this one work
Validation status
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Abstract

Reinforcement learning is a powerful paradigm for learning optimal policies from experimental data. However, to find optimal policies, most reinforcement learning algorithms explore all possible actions, which may be harmful for real-world systems. As a consequence, learning algorithms are rarely applied on safety-critical systems in the real world. In this paper, we present a learning algorithm that explicitly considers safety, defined in terms of stability guarantees. Specifically, we extend control-theoretic results on Lyapunov stability verification and show how to use statistical models of the dynamics to obtain high-performance control policies with provable stability certificates. Moreover, under additional regularity assumptions in terms of a Gaussian process prior, we prove that one can effectively and safely collect data in order to learn about the dynamics and thus both improve control performance and expand the safe region of the state space. In our experiments, we show how the resulting algorithm can safely optimize a neural network policy on a simulated inverted pendulum, without the pendulum ever falling down.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

The record

Venue
arXiv (Cornell University)
Topic
Adversarial Robustness in Machine Learning
Field
Computer Science
Canadian institutions
Funders
Natural Sciences and Engineering Research Council of CanadaSchweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
Keywords
Reinforcement learningInverted pendulumStability (learning theory)Computer scienceState spaceArtificial neural networkLyapunov functionGaussian processProcess (computing)State (computer science)Artificial intelligenceControl (management)Control theory (sociology)Machine learningGaussianAlgorithmMathematicsNonlinear system
Has abstract in OpenAlex
yes