gzilt-playground/rl

Experiments using Reinforcement Learning

A simple Reinforcement Learning Implementation

• Applied to the CartPole problem from OpenAi
• Based on Natural Evolution Strategies found at https://blog.openai.com/evolution-strategies/
• Also makes use of some of the code from http://kvfrans.com/simple-algoritms-for-solving-cartpole/

Dependency:

Requires the OpenAi gym

git clone https://github.com/openai/gym
cd gym
pip install -e . # minimal install