Robust estimators for functional logistic regression

GRACIELA BOENTE; MARINA VALDORA

Conferencia; 15th International Conference of the ERCIM WG on Computational and Methodological Statistics; 2022

The logistic regression model is widely used in data analysis. In many applied problems, the data are originated from phenomena that are better modeled by continuous functions than by finite-dimensional vectors. Functional logistic regression is a generalization of the logistic regression model that assumes that the covariates are functions while the responses are 0 or 1. We will present a robust proposal to estimate the slope under a functional logistic regression model which combines a dimension reduction with M-estimators. Through the results of a numerical study, we will illustrate the sensitivity of the classical estimator and the stability of the proposed method.