Supervised Machine Learning

Supervised learning in daily life

Whether working in industry or academia, game data scientists need to find patterns in data. This means extracting generalizable knowledge and insights, summarizing recurring patterns, and identifying similarities and differences from complex data. Pattern finding is central to game data science because it can sort out what is going on in voluminous, volatile, longitudinal, and sometimes sparse datasets.

To take an example, game companies using the freemium revenue model need to understand how players move from non-paying to paying players. To investigate the path of decisions and events that led a person to make the decision to invest cash in a game requires locating recurring patterns and trends among potentially millions of paths. This is where supervised machine learning methods, such as classification and regression analysis, can be incredibly useful. Such methods are in widespread use across both industry and academia.