Granular approach for evolving system modeling

Daniel Leite, Pyramo Costa, Fernando Gomide

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

21 Scopus citations

Abstract

In this paper we introduce a class of granular evolving system modeling approach within the framework of interval analysis. Our aim is to present an interval-based learning algorithm which develops both, granular and singular approximations of nonlinear nonstationary functions using singular data. The algorithm is capable of incrementally creating/adapting both model parameters and structure. These are key features in nonlinear systems modeling. In addition, interval analysis provides rigorous bounds on approximation errors, rounding errors, and on uncertainties in data propagated during computations. The learning algorithm is simple and particularly suited to process stream of data in real time. In this paper we focus on the foundations of the approach and on the details of the learning algorithm. An application concerning economic time series forecasting illustrates the usefulness and efficiency of the approach.

Original languageEnglish
Title of host publicationComputational Intelligence for Knowledge-Based Systems Design - 13th International Conference on Information Processing and Management of Uncertainty, IPMU 2010, Proceedings
Pages340-349
Number of pages10
DOIs
StatePublished - 2010
Externally publishedYes
Event13th International Conference on Information Processing and Management of Uncertainty, IPMU 2010 - Dortmund, Germany
Duration: 28 Jun 20102 Jul 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6178 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference13th International Conference on Information Processing and Management of Uncertainty, IPMU 2010
Country/TerritoryGermany
CityDortmund
Period28/06/102/07/10

Fingerprint

Dive into the research topics of 'Granular approach for evolving system modeling'. Together they form a unique fingerprint.

Cite this