An F# game like project using SDL (should be cross platform) that has two AI's fighting.

Programmed in a mostly functional manner but with mutability where necessary. Has Basic pathing, simple animations and projectiles, obstacle avoidance, planning and rendering. The main draw is the application of Reinforcement Learning (albeit not yet implmened) by the game bots and the use of an intelligence stat to affect learning.

The graphics are from one of my favorite childhood series. http://www.spiderwebsoftware.com/productsOld.html

##To Do

The code is mostly an expermiental playground, can RL learn to play a non-trivial game?

• Basic attack animations
• combination of elements to make enchanted weapons
• Reinforcement learning including actor-critic model, tracing and planning.