Designing unsupervised machine learning algorithms to explore user behaviour patterns for online games

网络游戏产生的海量时变和高维数据提供了新的 游戏行业和研究人员都有机会获得可操作的见解。这些数据可能 允许企业预测其收入，改善技术性能以增强用户的在线游戏 体验，根据用户的反馈或兴趣开发新游戏，最重要的是，预测 对某一游戏的未来有影响的玩家的行为。了解球员的行为是至关重要的 因为它既可以识别有价值的球员，也可以识别那些准备跳槽的球员。 这个项目的目标是调查如何分析游戏玩家产生的数据并 是否可以识别行为模式。

The massive amount of time-varying and high dimensional data generated from online games provides new opportunity to both the gaming industry and researchers to derive actionable insights. These data potentially allow businesses to predict their revenues, improve technical performances to enhance users' online gaming experience, develop new games based on users feedback or interests, and, most importantly, predict the behaviour of players with implications for future of a given game. Understanding players' behaviour is crucial as it allows the identification of both valuable players and also those players that are ready to churn. The objective of this project is to investigate how data generated by the game players can be analyzed and whether behavioural patterns can be identified.