Exploiting Clustering for Sports Data Analysis: A Study of Public and Real-World Datasets

Clustering as a data mining method has significant importance in data analysis. To achieve the goal of identifying prototypical features in sports data, this paper focuses on well-known clustering methods applied to publicly available data from the field of sports and activities, as well as a real-world dataset representing multi-domain measurements about professional athletes. Difficulties of the wide range of preprocessing methods as well as clustering methods are highlighted in this paper. In addition, the selected data sets are critically reviewed.