Component-Action Deep Q-Learning for Real-Time Strategy Game AI
In this thesis we will describe existing work in RTS AI and motivate our work adapting the deep Q-learning (DQN) RL algorithm to accommodate the multi- ...
Reinforcement learning applied to RTS gamesWe develop a set of techniques to mitigate the large branching factor problem. We apply such techniques in the scenario of a Real-Time Strategy (RTS) game, ... ELF: An Extensive, Lightweight and Flexible Research Platform for ...On ELF, we build a real-time strategy (RTS) game engine that includes three initial environments including. Mini-RTS, Capture the Flag and Tower Defense. Mini- ... Reinforcement Learning in RTS GamesThe common goal in the scenario used for the experiments, is letting multiple agents coop- eratively build a base in Balanced Annihilation. The fourth part, ...
Autres Cours:
