Tampere University of Technology

TUTCRIS Research Portal

Optimization procedure for predicting nonlinear time series based on a non-Gaussian noise model

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Details

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages540-549
Number of pages10
Volume4827 LNAI
DOIs
Publication statusPublished - 2007
Externally publishedYes
Publication typeA4 Article in a conference publication
Event6th Mexican International Conference on Artificial Intelligence, MICAI 2007 - Aguascalientes, Mexico
Duration: 4 Nov 200710 Nov 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4827 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Conference

Conference6th Mexican International Conference on Artificial Intelligence, MICAI 2007
CountryMexico
CityAguascalientes
Period4/11/0710/11/07

Abstract

In this article we investigate the influence of a Pareto-like noise model on the performance of an artificial neural network used to predict a nonlinear time series. A Pareto-like noise model is, in contrast to a Gaussian noise model, based on a power law distribution which has long tails compared to a Gaussian distribution. This allows for larger fluctuations in the deviation between predicted and observed values of the time series. We define an optimization procedure that minimizes the mean squared error of the predicted time series by maximizing the likelihood function based on the Pareto-like noise model. Numerical results for an artificial time series show that this noise model gives better results than a model based on Gaussian noise demonstrating that by allowing larger fluctuations the parameter space of the likelihood function can be search more efficiently. As a consequence, our results may indicate a more generic characteristics of optimization problems not restricted to problems from time series prediction.