C-LASSO estimator for generalized additive logistic regression based on B-Spline
[ 1 ] Instytut Inżynierii Bezpieczeństwa i Jakości, Wydział Inżynierii Zarządzania, Politechnika Poznańska | [ P ] employee
2019
chapter in monograph
english
EN Generalized Additive logistic Regression model (GALRM) is a very important nonparametric regression model. It can be used for binary classification or for predicting the certainty of a binary outcome by using generalized additive models, which known as modern techniques from statistical learning, and the penalized log-likelihood criterion. In our chapter, we develop an estimation problem for GALRM based on B-spline and Least Absolute Shrinkage and Selection Operator (LASSO), unlike the traditional solutions; we will express the LASSO problem as a conic quadratic optimization problem which is a well structured convex optimization program, and solve it great and very efficient interior points methods.
05.01.2019
173 - 190
20
20