A new approach for time series prediction using ensembles of ANFIS models

Authors:
Patricia Melin;Jesus Soto;Oscar Castillo;Jose Soria
Affiliations:
Tijuana Institute of Technology, P.O. Box 4207, Chula Vista, USA;Tijuana Institute of Technology, P.O. Box 4207, Chula Vista, USA;Tijuana Institute of Technology, P.O. Box 4207, Chula Vista, USA;Tijuana Institute of Technology, P.O. Box 4207, Chula Vista, USA
Venue:
Expert Systems with Applications: An International Journal
Year:
2012

Citing 3
Cited 4

Learning internal representations by error propagation

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Ensembling neural networks: many could be better than all

Artificial Intelligence
Fuzzy modeling using generalized neural networks and Kalman filter algorithm

AAAI'91 Proceedings of the ninth National conference on Artificial intelligence - Volume 2

An adaptive network-based fuzzy inference system (ANFIS) for the forecasting: The case of close price indices

Expert Systems with Applications: An International Journal
Construction of a new BRB based model for time series forecasting

Applied Soft Computing
Generalized dynamical fuzzy model for identification and prediction

Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology
A novel approach for change detection of remotely sensed images using semi-supervised multiple classifier system

Information Sciences: an International Journal

Quantified Score

Hi-index	12.05

Visualization

Abstract

This paper describes an architecture for ensembles of ANFIS (adaptive network based fuzzy inference system), with emphasis on its application to the prediction of chaotic time series, where the goal is to minimize the prediction error. The time series that we are considered are: the Mackey-Glass, Dow Jones and Mexican stock exchange. The methods used for the integration of the ensembles of ANFIS are: integrator by average and the integrator by weighted average. The performance obtained with this architecture overcomes several standard statistical approaches and neural network models reported in the literature by various researchers. In the experiments we changed the type of membership functions and the desired goal error, thereby increasing the complexity of the training.