Partial Least Squares Regression Estimation Of Nonorthogonal Problems

ABSTRACT In this project, Partial Least Square Regression was compared with Ordinary Least Square Regression (OLSR) to handle the problem of multicollinarity and small sample size on all Nigeria Insurance Company’s expenditure data. The prediction methods have been compared for efficiency through Root Mean Square Error (RMSE) and Mean Square Error (MSE). It is found that in this project Partial Least Square Regression (PLSR) provides better prediction as compared to the Ordinary Least Square Regression (OLSR).    TABLE OF CONTENTS Title page                                                     …