Focused Information Criteria for the Linear Hazard Regression Model

TL;DR

This paper introduces a focused information criterion for selecting the best model in the linear hazard regression framework, estimating mean squared error for each candidate model's cumulative hazard rate.

Contribution

It develops a novel model selection method based on a focused information criterion specifically for the linear hazard regression model.

Findings

The criterion effectively identifies models with minimal mean squared error.

Averaged versions of the criterion provide robust model selection.

The method offers an alternative to existing model selection techniques for nonparametric hazard models.

Abstract

The linear hazard regression model developed by Aalen is becoming an increasingly popular alternative to the Cox multiplicative hazard regression model. There are no methods in the literature for selecting among different candidate models of this nonparametric type, however. In the present paper a focused information criterion is developed for this task. The criterion works for each specified covariate vector, by estimating the mean squared error for each candidate model's estimate of the associated cumulative hazard rate; the finally selected model is the one with lowest estimated mean squared error. Averaged versions of the criterion are also developed.