Hide'n'Seek Reinforcement Learning

Introduction

This repo contains code for bachelor's project, which was about checking how the change of environment parameters affects agent's learning.

---

Folder structure

Analytics

In this folder you can find some analytics in Jupyter Notebook which solve the topic

django-compose

This folder contains whole project in one Docker container, incl. Flask, Redis, Celery and many more

hidenseek

Here you can find whole game written in PyGame. It allows you to test game mechanics without adapting it to the Gym OpenAI Environment guidelines.

progress_clips

This is an awesome archive showing the work progress of this project. Sadly it doesn't contain any code parts, only result videos. Starting from one square randomly moving and ending at fully functional Training Platform

---

Authors

IceDice

QuatZo

Woocashowski

Made as a Bachelor Project while studying at Silesian University of Technology, Gliwice, Poland