ESTIMATION OF SIMULTANEOUS SYSTEMS OF LINEAR EQUATIONS WITH NONLINEAR CONSTRAINTS AMONG THE COEFFICIENTS.
Abstract
The basic problem considered here is the estimation of the coefficients in a simultaneous system of linear equations. Maximum likelihood, linearized maximum likelihood, two-stage least-squares and three-stage least-squares methods have primarily been studied for the case of linear constraints among the coefficients. In this paper the case of nonlinear constraints is considered. It is shown that by linearizing the nonlinear constraints and using a three-stage least-squares procedure one can obtain similar large sample, and also small sigma, results. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 03, 1969
- Accession Number
- AD0684946
Entities
People
- Peter Ole Anderson
Organizations
- Stanford University