Abstract—In this paper, we use k-means, a basic type of clustering analysis, to analyze the data downloaded from the official site of NBA. We analyze the relevance between the number of clusters a team has and its winning percentage, the correlation coefficient is not high. However, we find that when the assuming number of clusters is ten, the relevance between the research items is the highest.
Index Terms—NBA, clustering analysis, correlation, statistics, winning percentage.
Mingzhe Xu is with Wuxi Big Bridge Academy, Jiangsu, China.
Junhui Gao is with the American and European International Study Center, Wuxi, China (Corresponding author: Junhui Gao; e-mail: jhgao68@163.com).
[PDF]
Cite: Mingzhe Xu and Junhui Gao, "Analysis of NBA Team Strength Using Players' Race Data Based on Clustering Method," International Journal of Social Science and Humanity vol. 7, no. 12, pp. 747-750, 2017.