Direct and recursive prediction of time series using mutual information selection

Authors:
Yongnan Ji;Jin Hao;Nima Reyhani;Amaury Lendasse
Affiliations:
Neural Network Research Centre, Helsinki University of Technology, Espoo, Finland;Neural Network Research Centre, Helsinki University of Technology, Espoo, Finland;Neural Network Research Centre, Helsinki University of Technology, Espoo, Finland;Neural Network Research Centre, Helsinki University of Technology, Espoo, Finland
Venue:
IWANN'05 Proceedings of the 8th international conference on Artificial Neural Networks: computational Intelligence and Bioinspired Systems
Year:
2005

Citing 7
Cited 16

Adaptive online learning algorithms for blind separation: maximum entropy and minimum mutual information

Neural Computation
Feature selection for high-dimensional genomic microarray data

ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
On Feature Selection: Learning with Exponentially Many Irrelevant Features as Training Examples

ICML '98 Proceedings of the Fifteenth International Conference on Machine Learning
Algorithms for Feature Selection: An Evaluation

ICPR '96 Proceedings of the 13th International Conference on Pattern Recognition - Volume 2
A study of cross-validation and bootstrap for accuracy estimation and model selection

IJCAI'95 Proceedings of the 14th international joint conference on Artificial intelligence - Volume 2
Training multilayer perceptron classifiers based on a modified support vector method

IEEE Transactions on Neural Networks
Input feature selection for classification problems

IEEE Transactions on Neural Networks

High-dimensional delay selection for regression models with mutual information and distance-to-diagonal criteria

Neurocomputing
Multigrid-based fuzzy systems for time series prediction: CATS competition

Neurocomputing
Recursive prediction for long term time series forecasting using advanced models

Neurocomputing
Sequential input selection algorithm for long-term prediction of time series

Neurocomputing
Multi-step ahead direct prediction for the machine condition prognosis using regression trees and neuro-fuzzy systems

Expert Systems with Applications: An International Journal
Long-term prediction of time series by combining direct and MIMO strategies

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Time series prediction using support vector machines: a survey

IEEE Computational Intelligence Magazine
Kernel methods applied to time series forecasting

IWANN'07 Proceedings of the 9th international work conference on Artificial neural networks
On incorporating seasonal information on recursive time series predictors

ICANN'07 Proceedings of the 17th international conference on Artificial neural networks
Approximate k-NN delta test minimization method using genetic algorithms: Application to time series

Neurocomputing
Multiple-output modeling for multi-step-ahead time series forecasting

Neurocomputing
Analysis of fast input selection: application in time series prediction

ICANN'06 Proceedings of the 16th international conference on Artificial Neural Networks - Volume Part II
Long-Term prediction of time series using state-space models

ICANN'06 Proceedings of the 16th international conference on Artificial Neural Networks - Volume Part II
Recursive diameter prediction for calculating merchantable volume of Eucalyptus clones without previous knowledge of total tree height using artificial neural networks

Applied Soft Computing
An integrated approach for healthcare planning over multi-dimensional data using long-term prediction

HIS'12 Proceedings of the First international conference on Health Information Science
Fast variable selection for memetracker phrases time series prediction

Proceedings of the 5th International Conference on PErvasive Technologies Related to Assistive Environments

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a comparison between direct and recursive prediction strategies. In order to perform the input selection, an approach based on mutual information is used. The mutual information is computed between all the possible input sets and the outputs. Least Squares Support Vector Machines are used as non-linear models to avoid local minima problems. Results are illustrated on the Poland electricity load benchmark and they show the superiority of the direct prediction strategy.