Bootstrap Aggregating Multivariate Adaptive Regression Splines (Bagging MARS) to Analyse the Lecturer Research Performance in Private University

Abstract

This study explores the use of a fairly simple nonparametric regression algorithm known as multivariate adaptive regression splines (MARS) which has the ability to approximate the relationship between the predictor and response variables. The main advantages of MARS are its capacity to produce simple, easy-to-interpret models, its ability to estimate the contributions of the predictor variables, and its computational efficiency. In this research, MARS is combined with one of nonparametric approach bootstrap aggregating (bagging). Bagging is used to improve the classification accuracy of the MARS method. This study is aimed to analyse lecturer research performance in private university using Bagging MARS algorithm. In modelling bagging MARS for lecturer research performance in a private university that there are three dominant influencing factors: 1) amount of research with cost an internal college granted by internal college with interest level of 86.20%; 2) number of publications of research results in national journals with interest level of 69.83%; and 3) number of speakers within national scientific meeting/ seminar with interest level of 63.34%. The accuracy of the MARS model classification is 84.615% with the classification error rate (APER) of 15.385%.