Features
Importance Samplings
- Importance sampling (IS)
- Per-decision Importance sampling (PDIS)
- Weighted Importance Sampling (WIS)
- Consistent Weighted Per-decision Importance sampling (CWPDIS)
Concentration Bounds
- Student's t
- Chernoff-Hoeffding's inequality
- Maurer & Pontil's Empirical Bernstein (MPeB) Inequality
- Eric & Philip's Monte Carlo m_alpha
Safe Offline RL Agent
- Seldonian Cross-Entropy method (CEM-Seldonian) agent