Bayesian MARS

被引：76

作者：

Denison, DGT

Mallick, BK

Smith, AFM

机构：

[1] Univ London Imperial Coll Sci Technol & Med, Dept Math, London SW7 2BZ, England

[2] Texas A&M Univ, Dept Stat, College Stn, TX 77843 USA

[3] Univ London Queen Mary & Westfield Coll, London E1 4NS, England

来源：

STATISTICS AND COMPUTING | 1998年 / 8卷 / 04期

基金：

英国工程与自然科学研究理事会;

关键词：

Bayesian methods; reversible jump Markov Chain Monte Carlo; multiple regression; multivariate adaptive regression splines;

D O I：

10.1023/A:1008824606259

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

A Bayesian approach to multivariate adaptive regression spline (MARS) fitting (Friedman, 1991) is proposed. This takes the form of a probability distribution over the space of possible MARS models which is explored using reversible jump Markov chain Monte Carlo methods (Green, 1995). The generated sample of MARS models produced is shown to have good predictive power when averaged and allows easy interpretation of the relative importance of predictors to the overall fit.

引用

页码：337 / 346

页数：10

共 29 条

[1]

[Anonymous], J AM STAT ASSOC

[2]

Becker R. A., 1988, NEW S LANGUAGE

[3]

BREIMAN L, 1985, J AM STAT ASSOC, V80, P580, DOI 10.2307/2288473

[4]

BRUNTZ SM, 1974, S ATM DIFF AIR POLL, P125

[5] Bayesian CART model search [J].

Chipman, HA ;

George, EI ;

McCulloch, RE .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1998, 93 (443) :935-948

[6]

CRAVEN P, 1979, NUMER MATH, V31, P317

[7]

Denison DGT, 1998, BIOMETRIKA, V85, P363

[8] Automatic Bayesian curve fitting [J].

Denison, DGT ;

Mallick, BK ;

Smith, AFM .

JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY, 1998, 60 :333-350

[9]

DENISON DGT, 1997, THESIS IMPERIAL COLL

[10] PROJECTION PURSUIT REGRESSION [J].

FRIEDMAN, JH ;

STUETZLE, W .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1981, 76 (376) :817-823

← 1 2 3 →