This repository contains a gym one player implementation of the classic boardgame Battleships and is intended to serve as a coding challenge for the seminar in deep reinforcement learning at ETH Zurich. The goal is to write a deep reinforcement learning algorithm that maximizes the given reward for destroying the battleships as fast as possible.
Clone repository
git clone https://github.com/OliverRichter/Coding_Challenge.gitInstall environment
cd Coding_Challenge/coding_challenge
pip install -e .Run random agent to see if installation is working
cd ../random_agent
python random_agent.py