An Examination on Applying Machine Learning in Game Industry

Abstract

Machine learning is one of the biggest trends in the current computer industry. However, the implementation of machine learning in commercial games is rare and limited to a certain minor feature in the game. That being said, there are a few implementations of machine learning in game industry that actually works. In this paper, we present our observation result regarding the implementation of machine learning in computer games. The observation focus on the architecture and methods used in the system, and the performance and advantage of using machine learning as part of the system. We go through several commercially available games that implement machine learning and analyze the architecture and performance of the system. Additionally, we also observe a few research related to machine learning that uses game as its environment by collecting several publications related to this area.