Frequently in linear regression analysis, the explanatory variables may be correlated among themselves and the data set may contain some observations that are outlying or extreme. In this article, we address an important problem: namely, accurately estimating the parameter in a linear model in the presence of outliers and multicollinear explanatory variables. The proposed solution is a combination of fuzzyweighted regression to mitigate the effect of the outliers and ridge regression to deal with the multicollinearity. We conduct a simulation study to evaluate the performance of the proposed estimator, and illustrate the methodology with an application.
Journal of Information & Optimization sciences 23(2), pp.259-271