Timezone: »

Security Analysis of Safe and Seldonian Reinforcement Learning Algorithms
Pinar Ozisik · Philip Thomas

Tue Dec 08 09:00 AM -- 11:00 AM (PST) @ Poster Session 1 #567

We analyze the extent to which existing methods rely on accurate training data for a specific class of reinforcement learning (RL) algorithms, known as Safe and Seldonian RL. We introduce a new measure of security to quantify the susceptibility to perturbations in training data by creating an attacker model that represents a worst-case analysis, and show that a couple of Seldonian RL methods are extremely sensitive to even a few data corruptions. We then introduce a new algorithm that is more robust against data corruptions, and demonstrate its usage in practice on some RL problems, including a grid-world and a diabetes treatment simulation.

Author Information

Pinar Ozisik (UMass Amherst)
Philip Thomas (University of Massachusetts Amherst)

More from the Same Authors