Machine Learning with Unreal help

Hi,

I have project where I need to create a simulation of an Air Hockey table. During game play I need to be able to get the x,y locations of the puck and paddles. These values will be feed into a Reinforcement Learning algorithm for training an agent to play the game. I have limited experience with unreal engine (although I am competent with c++). Would Unreal engine be overkill for this project? Ideally I would like to skip writing up collision physics e.t.c.

If it is possible could anyone give me any pointers?

Thanks

I recently published a Q – learning algorithm for reinforcement learning in Unreal Engine that can be used solve match to sample puzzles. This is a common experiment to assess instrumental learning skills in animals but I think it could be readily adapted to your Air Hockey AI. Links below, hope this helps.

Video here: https://www.youtube.com/watch?v=Wy7HDj2igPo
And the Marketplace link: Artificial Intelligence: Q-Learning in Blueprints - UE Marketplace