Nisa, Khoirin (2005) PENGKLASIFIKASIAN DATA MENGGUNAKAN ANALISIS DISKRIMINAN DAN ANALISIS GEROMBOL K-RATAAN. Jurnal Matematika Murni dan Terapan BKS-PTN Wilayah Barat, 3 (2). pp. 55-64. ISSN 1693-430X

[img] Text

Download (2MB)


Discriminant analysis and K-means cluster analysis are widely used statistical techniques for data classification. Here we compare the two methods by data simulation experiment. We generated data of the type most commonly assumed when using K-means clustering. The misclassifications rates and the R- squares (R2) of the two methods are discussed. The result shows that the discriminant analysis performs better than K-means clustering in misclassification rate, but worse in R2.

Item Type: Article
Subjects: H Social Sciences > HA Statistics
Depositing User: DR. KHOIRIN NISA
Date Deposited: 27 Mar 2018 04:43
Last Modified: 27 Mar 2018 04:43

Actions (login required)

View Item View Item