在眾多分群方法中，γ-SUP 演算法具有一些好的性質，使得它成為分群方法
當中很好的一項工具，然而此方法的初衷只是為了將相似度高的影像作分群，
並無法為每一群的中心估計作信賴區間的推論。這意味著 γ-SUP 演算法缺少了
樣本帶來的不確定性。在本文中，我們提出信賴區間的估計，並且希望能將這
套方法運用在 NBA 的資料上。過程中，我們以 γ-SUP 演算法為基礎，並且針對
NBA 的球員分群結果，建立每個群集的信賴區間，藉此了解每個群集的特色，
換句話說，我們想知道每群球員具有那些突出的能力，或是需要改善的缺點，
能夠提供教練在球員使用上的參考。另外根據我們的方法我們也發現了一些有
潛力成為明星球員的人，這項發現能夠實際的運用在球員自由市場上，提供球
隊管理層做為與球員簽定合約的依據。

For clustering methods, the γ-SUP algorithm possesses several favorable proper-
ties, making it a valuable tool in clustering. However, its original intention was merely
to cluster images with high similarity, without providing a confidence interval for each
cluster. It implies that the γ-SUP algorithm lacks evaluation for uncertainty from a sample. In this article, we provide a confidence interval and apply this method to NBA data.
Based on the γ-SUP algorithm, confidence intervals for clusters of NBA players are established to understand distinctive features between groups. In other words, we want to identify standout abilities or the needs of each player, which can provide managementdecisions in a free agent market.

1.Banfield, Raftery, D, J., & E, A. (1993). Model-based gaussian and non-gaussian clustering.
Biometrics, 803–821.
2.Chen, T.-L., Hsieh, D.-N., Hung, H., Tu, I.-P., Wu, P.-S., Wu, Y.-M., et al. (2014). γ-sup:
a clustering algorithm for cryo-electron microscopy images of asymmetric particles.
Journal of Structural Biology, 187(1), 1-9.
3.Chen, T.-L., & Shiu, S.-Y. (2007). A new clustering algorithm based on self-updating process.
JSM proceedings, statistical computing section, Salt Lake City, Utah, 2034–2038.
4.Chen, T.-L., & Shiu, S.-Y. (2012). Clustering by self-updating process. arXiv preprint
arxiv:1201.1979.
5.Cichocki, A., & Amari, S.-i. (2010). Families of alpha-beta-and gamma-divergences: Flex-
ible and robust measures of similarities. Entropy, 12(6), 1532–1568.
6.Eguchi, S., Komori, O., & Kato, S. (2011). Projective power entropy and maximum tsallis
entropy distributions. Entropy, 13(10), 1746–1764.
7.Fujisawa, H., & Eguchi, S. (2008). Robust parameter estimation with a small bias against
heavy contamination. Journal of Multivariate Analysis, 99(9), 2053–2081.
8.Hartigan, J. (1975). Clustering algorithms wiley. New York.
9.Lloyd, S. (1982). Least squares quantization in pcm. IEEE transactions on information
theory, 28(2), 129–137.
10.MacQueen, J., et al. (1967). Some methods for classification and analysis of multivariate ob-
servations. In Proceedings of the fifth berkeley symposium on mathematical statistics
and probability (Vol. 1, pp. 281–297).