Application of Data-Mining Techniques on Predictive Analysis

ABSTRACT: This paper compares five different predictive data-mining techniques on some data sets. The characteristics of these data are few predictor variables, many predictor variables, highly collinear variables, very redundant variables and presence of outliers. First, these data are pre-processed and prepared. In the first step after, preprocessing data, we must select a minimal number of variables that can completely predict the response variable. Different data-mining techniques is used in this research including: multiple linear regression MLR, principal component regression (PCR), an unsupervised technique based on the principal component analysis; ridge regression, the Partial Least Squares (PLS), and the Nonlinear Partial Least Squares (NLPLS). Each technique has different methods of usage; these different methods were used on each data set first and the best method in each technique was noted and used for global comparison with other techniques for the same data set.