Time series models discovery with similarity-based neuro-fuzzy networks and evolutionary algorithms

Author

Valdés, Julio J.

Author_Institution

Inst. for Inf. Technol., Nat. Res. Council of Canada, Montreal, Ont., Canada

Volume

3

fYear

2002

fDate

6/24/1905 12:00:00 AM

Firstpage

2345

Lastpage

2350

Abstract

The discovery of patterns of dependency in heterogeneous multivariate dynamic systems is approached with similarity-based neuro-fuzzy networks and evolutionary algorithms. The search space contains general autoregressive non-linear models representing the dependency structure of the process. Examples show that the proposed approach gives better results than the classical statistical one

Keywords

evolutionary computation; fuzzy systems; modelling; neural nets; pattern classification; search problems; time series; dependency structure; evolutionary algorithms; general autoregressive nonlinear models; heterogeneous multivariate dynamic systems; patterns discovery; search space; similarity-based neuro-fuzzy networks; time series models discovery; Councils; Evolutionary computation; Fuzzy neural networks; Fuzzy systems; Information technology; Neural networks; Predictive models; Robustness; Space technology; Time series analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 2002. IJCNN '02. Proceedings of the 2002 International Joint Conference on

Conference_Location

Honolulu, HI

ISSN

1098-7576

Print_ISBN

0-7803-7278-6

Type

conf

DOI

10.1109/IJCNN.2002.1007508

Filename

1007508