Choosing models in model-based clustering and discriminant analysis
Choosing models in model-based clustering and discriminant analysis
复制标题
DOI：
10.1080/00949659908811966
复制
发表时间：
1999-01-01
影响因子：
1.2
通讯作者：
Govaert, G
中科院分区：
文献类型：
--
作者：
Biernacki, C;Govaert, G
利用方差阵的特征值分解，Celeux和GovAert(1993)得到了大量和强大的基于高斯模型的聚类和判别分析模型。通过蒙特卡罗模拟，我们比较了选择这些模型的几种经典准则的性能：信息准则为AIC、贝叶斯准则BIG、分类准则为NEC和交叉验证。在集群环境中，信息标准和BIC优于分类标准。在判别分析的情况下，交叉验证表现出良好的性能，但信息准则和BIC也给出了令人满意的结果，而且到目前为止，计算时间要少得多。
Using an eigenvalue decomposition of variance matrices, Celeux and Govaert (1993) obtained numerous and powerful models for Gaussian model-based clustering and discriminant analysis. Through Monte Carlo simulations, we compare the performances of many classical criteria to select these models: information criteria as AIC, the Bayesian criterion BIG, classification criteria as NEC and cross-validation. In the clustering context, information criteria and BIC outperform the classification criteria. In the discriminant analysis context, cross-validation shows good performance but information criteria and BIC give satisfactory results as well with, by far, less time-computing.