Towards Safe Policy Improvement for Non-Stationary MDPs
Chandak, Yash, Jordan, Scott M., Theocharous, Georgios, White, Martha, Thomas, Philip S.
–arXiv.org Artificial Intelligence
Many real-world sequential decision-making problems involve critical systems with financial risks and human-life risks. While several works in the past have proposed methods that are safe for deployment, they assume that the underlying problem is stationary. However, many real-world problems of interest exhibit non-stationarity, and when stakes are high, the cost associated with a false stationarity assumption may be unacceptable. We take the first steps towards ensuring safety, with high confidence, for smoothly-varying non-stationary decision problems. Our proposed method extends a type of safe algorithm, called a Seldonian algorithm, through a synthesis of model-free reinforcement learning with time-series analysis. Safety is ensured using sequential hypothesis testing of a policy's forecasted performance, and confidence intervals are obtained using wild bootstrap.
arXiv.org Artificial Intelligence
Oct-23-2020
- Country:
- North America
- Canada (0.67)
- United States (1.00)
- North America
- Genre:
- Research Report > Experimental Study (0.46)
- Industry:
- Energy (0.67)
- Government > Regional Government
- North America Government > United States Government > FDA (0.46)
- Health & Medicine > Therapeutic Area
- Endocrinology > Diabetes (1.00)
- Information Technology (1.00)
- Technology: