EnemyUnknown Battlegrounds A trained surviver in the harsh environment!
Summary
As its name suggest, our project idea is inspired by the most polular vedio game PlayerUnknown’s Battlegrouds. The goal of this project is to have an agent to survive in the 40x40 flat ground with three enemies(zombies), two pigs, and ten traps spawned randomly in the environment. The agent should survive as long as possible to escape from the acttack of zombies or to kill pigs to increase its health points.
Links
Resources Used:
We mainly used the class notes and textbook for our reference material. Below are some additional resources for us to learn and improve our algorithm.
1.The resource used for deep q learning tutorial:https://sergioskar.github.io/Deep_Q_Learning/
2.The simple reinforcement learning tutorial: https://medium.com/emergent-future/simple-reinforcement-learning-with-tensorflow-part-0-q-learning-with-tables-and-neural-networks-d195264329d0