Deep Q-learning using redundant outputs in visual doom

摘要

Recently, there is a growing interest in applying deep learning in game AI domain. Among them, deep reinforcement learning is the most famous in game AI communities. In this paper, we propose to use redundant outputs in order to adapt training progress in deep reinforcement learning. We compare our method with general ε-greedy in ViZDoom platform. Since AI player should select an action only based on visual input in the platform, it is suitable for deep reinforcement learning research. Experimental results show that our proposed method archives competitive performance to ε-greedy without parameter tuning.