Maximum Likelihood Principle and Model Selection when the True Model is Unspecified.
Abstract
Suppose independent observations come from an unspecified distribution. Then we consider the maximum likelihood based on a specified parametric family by which we can approximate the true distribution well. We examine the asymptotic properties of the quasi-maximum likelihood estimate and of the quasi-maximum likelihood. These results will be applied to model selection problem. Keywords: AIC, BIC, Consistency. Law of iterated logarithm MLE, Regularity conditions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1987
- Accession Number
- ADA186027
Entities
People
- Ryuei Nishii
Organizations
- University of Pittsburgh