Performance Prediction of Players in Sports League Matches

The objective of this article is to discover the better performing team in the Hockey India League (HIL) for the purpose of formation of the winning team based on cluster analysis of their past performance by using the machine learning techniques. Two most prevalent machine learning techniques k-means and fuzzy clustering have been used respectively to predict the better performing player. The results of the two techniques proposed were compared and were found nearly identical. The complexity of initializing K-means clustering technique is resolved by using MacQueen algorithm. The results obtained from Hockey Indian League Goal statistics dataset were used to detect n-clusters to handle the imprecise and ambiguous result. Finally, this article proposed a K-Means clustering technique which provides efficient and accurate data analysis in the field of data mining.