Skip to content

Reinforcement Learning: Cartpole Balancing with a DQN Agent

License

Notifications You must be signed in to change notification settings

daniels-kb/openai-dqn-cartpole

Repository files navigation

Project-CartpoleDQN

COMP532 Group Assignment 2 - Python Cartpole DQN agent

In this assignment we aimed to design and implement a deep RL agent for a video game using OpenAI Gym, an open source interface to reinforcement learning tasks.

The game of choice was Cartpole, where an agent had to learn to balance a pole on top of a cart running on frictionless tracks. More in report pdf. One can see a visual representation in the gif