Maximum trimmed likelihood estimator for multivariate mixed continuous and categorical data

Authors:
Tsung-Chi Cheng;Atanu Biswas
Affiliations:
Department of Statistics, National Chengchi University, 64 ZhihNan Road, Section 2, Taipei 11605, Taiwan;Applied Statistics Unit, Indian Statistical Institute, 203 B.T. Road, Kolkata 700 108, India
Venue:
Computational Statistics & Data Analysis
Year:
2008

Citing 8
Cited 2

Statistical analysis with missing data

Statistical analysis with missing data
Robust regression and outlier detection

Robust regression and outlier detection
Generalization of the Mahalanobis distance in the mixed case

Journal of Multivariate Analysis
Maximum trimmed likelihood estimators: a unified approach, examples, and algorithms

Computational Statistics & Data Analysis
A fast algorithm for the minimum covariance determinant estimator

Technometrics
Influence function and efficiency of the minimum covariance determinant scatter matrix estimator

Journal of Multivariate Analysis
A generalized Mahalanobis distance for mixed data

Journal of Multivariate Analysis
Robust regression diagnostics with data transformations

Computational Statistics & Data Analysis

Robust diagnostics for the heteroscedastic regression model

Computational Statistics & Data Analysis
Pointwise probability reinforcements for robust statistical inference

Neural Networks

Quantified Score

Hi-index	0.03

Visualization

Abstract

In this article, we apply the maximum trimmed likelihood (MTL) approach [Hadi, A.S., Luceno, A., 1997. Maximum trimmed likelihood estimators: a unified approach, examples, and algorithms. Comput. Statist. Data Anal. 25, 251-272] to obtain the robust estimators of multivariate location and shape, especially for data mixed with continuous and categorical variables. The forward search algorithm [Atkinson, A.C., 1994. Fast very robust methods for the detection of multiple outliers. J. Amer. Statist. Assoc. 89, 1329-1339] is adapted to compute the proposed MTL estimates. A simulation study shows that the proposed estimator outperforms the classical maximum likelihood estimator when outliers exist in data. Real data sets are also used to illustrate the method and results of the detection of the outliers.