2006•Encyclopedia of EnvironmetricsRequires access

Multivariate Adaptive Regression Splines

Charles Kooperberg

Abstract

Abstract The multivariate adaptive regression splines (MARS) algorithm was developed by Friedman as a nonparametric method for multiple regression. The MARS algorithm uses adaptively selected spline functions. Because of its adaptivity MARS is often able to model functions that express high variation in one region of the predictor space and are smoother in other parts.

About this research paper

What this paper is about

Why it matters

OpenAlex reports 129 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Key concepts: Multivariate adaptive regression splines, Nonparametric regression, Multivariate statistics, Mars Exploration Program, Spline (mechanical), Regression, Nonparametric statistics, Bayesian multivariate linear regression