Research of Principal Component Analysis in Software Metrics
Hao Ke-gang · Computer Technology and Development · 2006
In order to decrease the dimension of mass correlated dataset during data analysis process of software metrics,in the paper,introduce the principal component analysis(PCA) to this period.On the base of keeping as more as possible of the original features of the dataset,successfully lower the dimension of the raw data,and get the main factor which affect the object that we are measured.In this way,make the meaning of measure data clearer and more accurate,and make the whole metrics process more effective.