A Case Study on the RCMD Method and Fuzzy C-Regression Models for Mining Regression Classes
Jianghong Ma, Guojun Wang · 2008
It is well known that the classical regression analysis, especially parametric regression analysis, is one of the important methods of extracting information from data sets. A family of regression functions, called fuzzy c-regression models (FCRM), has been presented which can be used to characterize the linear relationship to certain types of mixed data. More generally, an effective and robust method, coined regression class mixture decomposition (RCMD), has also been proposed for the mining of regression classes in large data sets. In this paper, we focus on a special case of regression class mixture models, switching regression models, and adopt the RCMD method and the FCRM method in a real example and a simulation experiment. It is shown that the RCMD method has some special advantages over some traditional methods of data analysis and these two methods give almost consistent estimation results in switching regression models.