Efficient estimation of multidimensional regression model using multilayer perceptrons

Authors:
Joseph Rynkiewicz
Affiliations:
Université Paris I, SAMOS/MATISSE, 90 rue de Tolbiac, Paris, France
Venue:
Neurocomputing
Year:
2006

Citing 6
Cited 1

Nonlinear statistical models

Nonlinear statistical models
Original Contribution: Uniqueness of the weights for minimal feedforward nets with a given input-output map

Neural Networks
A regularity condition of the information matrix of a multilayer perceptron network

Neural Networks
System identification (2nd ed.): theory for the user

System identification (2nd ed.): theory for the user
Using neural networks to model conditional multivariate densities

Neural Computation
Neural modeling for time series: A statistical stepwise method for weight elimination

IEEE Transactions on Neural Networks

Signal denoising in engineering problems through the minimum gradient method

Neurocomputing

Quantified Score

Hi-index	0.01

Visualization

Abstract

This work concerns the estimation of multidimensional non-linear regression models using multilayer perceptrons (MLPs). The main problem with such models is that we need to know the covariance matrix of the noise to get an optimal estimator. However, we show in this paper that if we choose as the cost function the logarithm of the determinant of the empirical error covariance matrix, then we get an asymptotically optimal estimator. Moreover, under suitable assumptions, we show that this cost function leads to a very simple asymptotic law for testing the number of parameters of an identifiable MLP. Numerical experiments confirm the theoretical results.