/stochastic_policy_search

Implementation of three stochastic policy search algorithms for RL

Primary LanguageJupyter Notebook

Hill Climbing

This repository contains the implementation of the following stochastic policy search algorithms:

  • Vanilla hill climbing
  • Hill climbing with simulated annealing
  • Steepest ascent hill climbing with adaptive noise scaling

They are all applied to the OpenAi Gym Cartpole-v0 environment, for simplicity.