https://github.com/jedrazb/rl-grid-world
Reinforcement learning agent which finds a path to the goal in a grid world. This exercise was done as a coursework for course C424 at Imperial College London.
https://github.com/jedrazb/rl-grid-world
deep-q-learning deep-q-network grid-world reinforcement-agents reinforcement-learning reinforcement-learning-agent
Last synced: 3 months ago
JSON representation
Reinforcement learning agent which finds a path to the goal in a grid world. This exercise was done as a coursework for course C424 at Imperial College London.
- Host: GitHub
- URL: https://github.com/jedrazb/rl-grid-world
- Owner: jedrazb
- Created: 2020-01-08T19:34:24.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2020-01-08T20:13:55.000Z (over 6 years ago)
- Last Synced: 2025-02-09T18:49:41.747Z (over 1 year ago)
- Topics: deep-q-learning, deep-q-network, grid-world, reinforcement-agents, reinforcement-learning, reinforcement-learning-agent
- Language: Python
- Homepage:
- Size: 12.7 KB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Deep Q-learning agent in the grid world
Deep Q-learning agent which finds a path to the goal in a grid world. This exercise was done as a coursework for course C424 at Imperial College London.
--------
Dependencies: `numpy`, `cv2`, `torch`
To start the training run:
```
python train_and_test.py
```
-------
Following techniques were used:
* Deep Q-network (created in PyTorch)
* Models with continous (`agent_radians.py`) and discrete actions (`agent.py`)
* Prioritised experience replay buffer
* Epsilon greedy policy
* Target network
* Sampling using Cross Entropy Method