Skip to content

Releases: thakorneyp11/reinforcement-learning

v1-2023.16.0

23 Apr 21:11
@thakorneyp11 thakorneyp11
v1-2023.16.0
7c51ee6
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Verified
Learn about vigilant mode.
Compare
Choose a tag to compare
v1-2023.16.0 Latest
Latest

Overview

  • initial version on implementing reinforcement learning approaches from scratch

Current Features

Implemented Algorithms

  • Cross Entropy method
  • Monte Carlo: MC Prediction, MC Control
  • Temporal Difference: N-step SARSA, SARSAmax, Expected SARSA
  • Temporal Difference: Double Q-Learning
  • Deep Q Network (DQN)

Implemented Tools

  • Customized Gridworld environment
  • ONNX model conversion and usage script

Newly Added Features

  • Feat: add Monte Carlo algorithm and custom Gridworld environment (PR #1)
  • Feat: add Temporal Difference algorithm (PR #2)
  • Feat: add algorithm conversion to ONNX format (PR #3)
  • Feat: add Cross Entropy method on Cartpole environment (PR #4)
  • Feat: add Double Q-Learning on Taxi environment (PR #5)
  • Feat: add DQN algorithm on MountainCar environment (PR #7)
Assets 2
Loading