1972BiometrikaRequires access

Maximum Likelihood Estimation of the General Nonlinear Functional Relationship with Replicated Observations and Correlated Errors

G. R. Dolby, S. Lipton

Open publisher page 1 citations

Abstract

The problem considered is that of estimating a p-parameter functional relationship η = η(ξ;α1, …, αp given replicated observations at each of n unknown values of the independent variable ξ. The errors of observation at (ξ, η) are assumed to have a bivariate normal distribution involving parameters σ2i, pi and r2i and to be independent of those at any other true point. An iterative method for obtaining maximum likelihood estimates of the structural and incidental parameters is proposed. Closed form expressions in terms of the ξi are obtained for the asymptotic covariance matrix of estimates of the structural parameters αv and the scale incidental parameters σv, pi and τ2i. An illustrative application to simulated data and a discussion of convergence are included.

About this research paper

What this paper is about

The problem considered is that of estimating a p-parameter functional relationship η = η(ξ;α1, …, αp given replicated observations at each of n unknown values of the independent variable ξ. The errors of observation at (ξ, η) are assumed to have a bivariate normal distribution involving parameters σ2i, pi and r2i and to be independent of those at any other true point. An iterative method for obtaining maximum likelihood estimates of the structural and incidental parameters is proposed. Closed form expressions in terms of the ξi are obtained for the asymptotic covariance matrix of estimates of the structural parameters αv and the scale incidental parameters σv, pi and τ2i. An illustrative application to simulated data and a discussion of convergence are included.

Why it matters

OpenAlex reports 1 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

The problem considered is that of estimating a p-parameter functional relationship η = η(ξ;α1, …, αp given replicated observations at each of n unknown values of the independent variable ξ. The errors of observation at (ξ, η) are assumed to have a bivariate normal distribution involving parameters σ2i, pi and r2i and to be independent of those at any other true point. An iterative method for obtaining maximum likelihood estimates of the structural and incidental parameters is proposed. Closed form expressions in terms of the ξi are obtained for the asymptotic covariance matrix of estimates of the structural parameters αv and the scale incidental parameters σv, pi and τ2i. An illustrative application to simulated data and a discussion of convergence are included.

Key concepts: Mathematics, Bivariate analysis, Covariance, Covariance matrix, Statistics, Maximum likelihood, Convergence (economics), Applied mathematics

Related papers

Back to paper searchBrowse research topicsOriginal source
Maximum Likelihood Estimation of the General Nonlinear Functional Relationship with Replicated Observations and Correlated Errors — Research Paper | ScholarLens