Adaptive Sampling using Stein's Discrepancy

Feb 23, 2022
This project is WIP (due to being a Master’s thesis), Code and Report will be released in April.

Project Description

Investigation into the usage of different sampling methods based on Stein’s discrepancy in the context of Curriculum Learning.

Implemented both a gradient-based approach (Stein’s Variational Gradient Descent1) and a direct minimization method (Stein Points2) in jax and evaluated them on different RL baselines.

  1. Liu et al., Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm ↩︎

  2. Chen et al., Stein Points ↩︎

reinforcement-learning sampling jax research
Heiko Carrasco
Heiko Carrasco
Software Engineer