Subject Grouping Using Mahalanobis Distance and PCA

Melda Putri Boni
2018 International Journal for Research in Applied Science and Engineering Technology  
In multivariate analysis PCA can be used to reduce data so that data that initially has many variables will produce a few variables making it easier to do the grouping, and with the PCA then the multicolloniarity data can be overcome. The use of PCA is the Robust PCA (ROBPCA) so that data is not easily influenced by outliers. With PCA reductions, clustering with cluster analysis using Mahalanobis Distance similarity will result in more optimal grouping, since the outliers have been overcome
more » ... e been overcome with Robust using Minimum Covariance Determinant (MCD) and Minimum Volume Ellipsoid (MVE).
doi:10.22214/ijraset.2018.4415 fatcat:caulhi7ay5bifnmeo4cjvmyaem