On Robustifying the Fishers Discriminant Function using L Estimators

Kennet G. Cuarteros, Emily Amor A. Balase


Multivariate data can be classified into different groups. One useful statistical tool for classification is discriminant analysis whose major objective is to classify data into different populations based on a training sample. The problem arises when the data contain outliers which greatly affect the classification performance. Some studies used robust L-estimators such as median, truncated mean, trimean, and winsorized mean, yielding a robust version of Fishers Discriminant Function. In this study, the total probability of misclassification is computed through a simulation experiment using MATLAB to examine the behavior of the L-estimators. Relative efficiencies are determined to compare the efficiency of the estimators. Results showed that, when using the robust L-estimators, the classification performance of the discriminant rules improved, and among the estimators, median is appropriate for classifying observations but the classification efficiency is limited. L-estimators outperformed the classical in terms of the relative efficiency. Among the L-estimators, winsorized mean is more stable in terms of classification efficiency.


discriminant analysis, robust estimators, L-estimators, total probability of misclassification, relative efficiency

Full Text:



Altman, E., (1968). Financial ratios, discriminant analysis and the prediction of corporate bankruptcy.Journal of Finance 23 (4), 589-609.

Beaver, W., (1968). Financial ratios predictors of failure. Journal of Accounting Research 4, 71-111.Balase, E., Padua, R., 2006. Robust L-Discriminant Analysis: Asymptotics, Simulation and Monte Carlo. The Philippine Statistician 55: 5564.

Chellappa, R., Etemad, K., (1997). Discriminant analysis for recognition of human face images, Journal of the Optical Society of America A 14, 1724-1733.

Chapman, D.W., Hutcheson, S.M., (1982). Attrition from Teaching Careers: A Discriminant Analysis. American Educational Research Journal 19: 93-105.

Croux, C., Haesbroeck, G. (1999). Empirical Influence Function for Robust Principal component analysis.

Croux, C., Dehon, C. 2001. Robust linear discriminant analysis using S-estimators. Canad. J. Statist. 29: 473492.

Cuarteros, K., Puerte, R., 2017. Classifying Students Engagement in

Computer Games using Linear Discriminant Analysis. International Journal of Science and Research,6 (7): 715-722.

Dudoit, S., et al. (2002). Comparison of Discrimination Methods for the Classification of Tumors Using Gene Expression Data. Journal of the American Statistical Association,97(457): 77-87.

Emel, A.B., et al. (2003). A Credit Scoring Approach for the Commercial Banking Sector. Socio-Economic Panning Sciences 37:103-123.

Gomez, J.C., Moens, M.F. (2010). Using Biased Discriminant Analysis for Email Filtering. KES'10 Proceedings of the 14th International Conference on Knowledge-based and Intelligent Information and Engineering Systems:566-574.

Hubert, M., Van Driessen, K., (2004). Fast and Robust Discriminant Analysis. Computational Statistics & Data Analysis 45, 301320.

Hubert, M., Engelen, S., (2004). Robust PCA and classification in biosciences. Bioinformatics 20: 17281736.

Kennedy, JW., et al. (1980). Multivariate discriminant analysis of the clinical and angiographic predictors of operative mortality from the Collaborative Study in Coronary Artery Surgery (CASS). The Journal of Thoracic and Cardiovascular Surgery 80: 876-887.

Kumar, N., Andreou, A.G. (1998). Heteroscedastic Discriminant Analysis and Reduced Rank HMMs for Improved Speech Recognition. Speech Communication,25.

Staudte R., and S. Sheather (1990). Robust Estimation and Testing (Wiley Series, 1990). R. Stiegler. L-Estimation of Location Parameter: Asymptotics and Monte Carlo. Journal of the American Statistical Association, 1969


  • There are currently no refbacks.