Subscribe Now Subscribe Today
Science Alert
Curve Top
Asian Journal of Mathematics & Statistics
  Year: 2010 | Volume: 3 | Issue: 2 | Page No.: 93-101
DOI: 10.3923/ajms.2010.93.101
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail

Estimating Multinomial Logit Model with Multicollinear Data

I. Camminatiello and A. Lucadamo

The multinomial logit model is used to study the dependence relationship between a categorical response variable with more than two categories and a set of explicative variables. In presence of multicollinearity, the estimation of the multinomial logit model parameters becomes inaccurate. To solve this problem we develop an extension of principal component logistic regression. Finally a simulation study illustrates the advantages of the method.
PDF Fulltext XML References Citation Report Citation
  •    An Alternative Multicollinearity Approach in Solving Multiple Regression Problem
  •    Central Limit Theorem for the Sum of a Random Number of Dependent Random Variables
How to cite this article:

I. Camminatiello and A. Lucadamo, 2010. Estimating Multinomial Logit Model with Multicollinear Data. Asian Journal of Mathematics & Statistics, 3: 93-101.

DOI: 10.3923/ajms.2010.93.101






Curve Bottom