/gradient-ascent-stochastic-policy-learning

Open AI Cartpole environment gradient ascent

Primary LanguageJupyter Notebook

gradient-ascent-cartpoleEnv

Open AI Cartpole environment gradient ascent

Implementation of gradient ascent for policy learning in DRL

Includes

  • Cross entropy model
  • Adoptive noise scaling

Other methods for gradient ascent (TO DO)

  • Steepest ascent (Hill climbing)
  • Stimulated annealing