Abstract : K-means is a classic, the division of the clustering algorithm, apply to the classification of the globular data. According to the initial clustering center, this paper comprehensive consideration the characteristics of various Hierarchical cluster algorithms and choose the appropriate Hierarchical cluster algorithm to improve K-means, and combined with Hainan Green Tangerine Peel cluster analysis of data which is compared experiments. The results indicate that the improved algorithm have increasing the distance between classes with each others, get a stable of cluster results and better implementation data mining. Finally to summary the two algorithms and the further research direction.
Hongxiu Li; Matti Mäntymäki; Xianfeng Zhang. 13th Conference on e-Business, e-Services and e-Society (I3E), Nov 2014, Sanya, China. Springer, IFIP Advances in Information and Communication Technology, AICT-445, pp.144-150, 2014, Digital Services and Information Intelligence. 〈10.1007/978-3-662-45526-5_14〉
https://hal.inria.fr/hal-01342139
Contributeur : Hal Ifip
<>
Soumis le : mardi 5 juillet 2016 - 14:39:06
Dernière modification le : mercredi 6 juillet 2016 - 10:09:17
Ying Luo, Haiyan Fu. Modified K-means Algorithm for Clustering Analysis of Hainan Green Tangerine Peel. Hongxiu Li; Matti Mäntymäki; Xianfeng Zhang. 13th Conference on e-Business, e-Services and e-Society (I3E), Nov 2014, Sanya, China. Springer, IFIP Advances in Information and Communication Technology, AICT-445, pp.144-150, 2014, Digital Services and Information Intelligence. 〈10.1007/978-3-662-45526-5_14〉. 〈hal-01342139〉
