Global Safe Sequential Learning via Efficient Knowledge Transfer

TMLR Paper2172 Authors

Abstract:

Sequential learning methods such as active learning and Bayesian optimization select the most informative data to learn about a task.
In many medical or engineering applications, the data selection is constrained by a priori unknown safety conditions.
A promissing line of safe learning methods utilize Gaussian processes (GPs) to model the safety probability and perform data selection in areas with high safety confidence.
However, accurate safety modeling requires prior knowledge or consumes data.
In addition, the safety confidence centers around the given observations which leads to local exploration.
As transferable source knowledge is often available in safety critical experiments, we propose to consider transfer safe sequential learning to accelerate the learning of safety.
We further consider a pre-computation of source components to reduce the additional computational load that is introduced by incorporating source data.
In this paper, we theoretically analyze the maximum explorable safe regions of conventional safe learning methods.
Furthermore, we empirically demonstrate that our approach 1) learns a task with lower data consumption, 2) globally explores multiple disjoint safe regions under guidance of the source knowledge, and 3) operates with computation comparable to conventional safe learning methods.

Submission Length: Regular submission (no more than 12 pages of main content)

Changes Since Last Submission:

We list the changes below. For easy readability, we further attach an edit track pdf in supplementary (anonymized, latexdiff tool).

- related work updated (Sergeyev et al. (2020))

- notation significantly changed for clarity (summarized in table 1)

- remark 3.2, 4.2 added

- new experiments added to section 5 (and the attached numerical details in appendix D)

- new experiments also mean table 3, figure 3 bottom, figure 4 top are updated

- ablation studies added to appendix E, figure 8 & 9

- move one figure from appendix to main (page 6)

- conclusion, add one sentence (last sentence)

Assigned Action Editor: ~Branislav_Kveton1

Submission Number: 2172