Melik, Hameedah Naeem and Ahad, Nor Aishah and Syed Yahaya, Sharipah Soaad (2018) Modified one-step M-estimator with robust scale estimator for multivariate data. Journal of Engineering and Applied Sciences, 13 (24). pp. 10396-10400. ISSN 1816-949X

PDF Restricted to Registered users only Download (891kB) | Request a copy

Abstract

The Modified One-step M-estimator (MOM) is a highly efficient robust estimator for classifying multivariate data. Generally, robust estimators came into existence as a solution to the inability of classical Linear Discriminant Analysis (LDA) to perform optimally in the presence of outliers. Thus, to solve this shortcoming, the robust MOM estimator is integrated with a highly robust scale estimator, Qn, in the trimming criterion of MOM. This introduces a new robust approach termed RLDAMQ for handling outliers encountered in multivariate data. The results show the superiority of RLDAMQ over the classical LDA and previously existing robust method in literature in terms of misclassification error evaluated through simulated data.

Item Type:	Article
Uncontrolled Keywords:	Modified One-Step M-Estimator, Robust.Q..multivariate data, trimming,criterion, encountered
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	School of Quantitative Sciences
Depositing User:	Mrs. Norazmilah Yaakub
Date Deposited:	01 Oct 2020 03:06
Last Modified:	01 Oct 2020 03:06
URI:	https://repo.uum.edu.my/id/eprint/27550

Actions (login required)

View Item