Data mining Application of Data Reduction and Clustering Domain of Textile Database

Abstract

This research paper attempts to identify the textile data structure and hidden pattern of original database with certain important parameters. The main objectives of this study are to identify the first n number of factors that explained over the study period. Initially factor analysis is performed to extract factor scores. Principal extraction is performed through Data mining package with sixteen textile fabrics parameters. Factor extraction is aimed to uncover the intrinsic pattern among the textile parameters considered and an important point of factor analysis is to extract factor scores for further investigation. Thus, factor analysis consistently resulted in three factors for the whole datasets. The amount of total variation explained is over 75 percent in factor analysis with varimax rotation. The factor loadings or factor structure matrix with unassociated rotation methods are not always easy to interpret. The nonhierarchical k-mean clustering is also used to identify meaningful cluster based on their parameter means of original database.