UE4 Reinforcement Learning Component

Release: 2020

The reinforcement learning component is a plugin for Unreal Engine 4 that provides game developers an accessible way to incorporate machine learning within their game AIs.

  • Dynamic creation of Q-tables that matches the defined states and actions
  • Monitoring tools help finding and tweaking the correct parameters to get the best possible results
  • Switch the AI to play mode and declare a difficulty function to dynamically make the AI more and less clever to provide the best challenge for players of all different experience levels

Technical Realization:

The reinforcement component is coded entirely in UE4 C++.

Images: