Estimating the Dimension of a Model
The problem of selecting one of a number of models of different dimensions is treated by finding its Bayes solution, and evaluating the leading terms of its asymptotic expansion. These terms are a valid large-sample criterion beyond the Bayesian context, since they do not depend on the a priori distribution.
Permanent link to this document: http://projecteuclid.org/euclid.aos/1176344136
Digital Object Identifier: doi:10.1214/aos/1176344136
Mathematical Reviews number (MathSciNet): MR468014
Zentralblatt MATH identifier: 0379.62005