Batch learning competitive associative net and its application to time series prediction

Kurogi, Shuichi; Ueno, Tomoya; Sawa, Miho

doi:10.1109/ijcnn.2004.1380195

Cited by 11 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, β is the undetermined multiplier for the condition (10) in the Lagrange method. Moreover, β derived from (9)- (11) has no relation with n, while β = β n derived from (6) increases exponentially with the increase of n.…”

Section: A Bayesian Bagging Predictionmentioning

confidence: 95%

See 1 more Smart Citation

Improving generalization performance of bagging ensemble via Bayesian approach

Kurogi

Harashima

2009

2009 IEEE International Symposium on Computational Intelligence in Robotics and Automation - (CIRA)

View full text Add to dashboard Cite

This paper describes a method for improving the generalization performance of bagging ensemble by means of using Bayesian approach. We examine the Bayesian prediction using bagging leaning machines for regression problems, and show a method to reduce the generalization loss defined by the square error of the prediction for test data. We examine and validate the effectiveness via numerical experiments using the CAN2s as learning machines, where the CAN2 is a neural net for learning efficient piecewise linear approximation of nonlinear functions.

show abstract

Section: A Bayesian Bagging Predictionmentioning

confidence: 95%

“…Note that we have developed an efficient batch learning method (see [11] for details), which we have used in this application.…”

Section: Appendixmentioning

confidence: 99%

Improving generalization performance of bagging ensemble via Bayesian approach

Kurogi

Harashima

2009

2009 IEEE International Symposium on Computational Intelligence in Robotics and Automation - (CIRA)

View full text Add to dashboard Cite

show abstract

“…Note that we have developed an efficient batch learning method (see [6] for details), and we use it in the present competition. The method consists of iterations of (1) competitive learning based on a gradient method, (2) associative learning employing recursive least squares, and (3) reinitialization of units based on an "asymptotic optimality" criterion (see [4]) for overcoming local minima problems of the gradient method.…”

Section: Can2 and The Bagging 1) Assumptions On The Given Dataset:mentioning

confidence: 99%

“…The method consists of iterations of (1) competitive learning based on a gradient method, (2) associative learning employing recursive least squares, and (3) reinitialization of units based on an "asymptotic optimality" criterion (see [4]) for overcoming local minima problems of the gradient method. We have used the same parameter values as for the function approximation problems shown in [6], except the number of units involved in the CAN2 which is tuned so that the prediction achieves smaller SMAPE for the validation periods (see Section II-B and Section III).…”

Section: Can2 and The Bagging 1) Assumptions On The Given Dataset:mentioning

confidence: 99%

Forecasting Using First-Order Difference of Time Series and Bagging of Competitive Associative Nets

Kurogi

Koyama

Tanaka

et al. 2007

2007 International Joint Conference on Neural Networks

View full text Add to dashboard Cite

Abstract-This article describes our method used for the 2007 Forecasting Competition for Neural Networks and Computational Intelligence. We have employed the first-order difference of time series for dealing with the seasonality of the monthly data. Since the differencing removes the trend of time series, we have developed a method to estimate the trend. Moreover, we have used the bagging of competitive associative net called CAN2 as a learning predictor, where the CAN2 is for learning an efficient piecewise linear approximation of a nonlinear function, and the bagging for reducing the variance of the prediction.

show abstract

“…In order to optimize w l i and M l i for i ∈ I l , we have developed an efficient batch learning method (see [7], [8] for details), which we have used at the Competition. For focusing on new method for the Competition, we here omit to describe the learning method.…”

Section: A Prediction By the Can2 And Can2 Ensemblementioning

confidence: 99%

Ensemble of Competitive Associative Nets and Multiple K-fold Cross-Validation for Estimating Predictive Uncertainty in Environmental Modelling

Kurogi¹,

Kuwahara²,

Tanaka³

The 2006 IEEE International Joint Conference on Neural Network Proceedings

Self Cite

View full text Add to dashboard Cite

This article describes the method which we have used for the Predictive Uncertainty in Environmental ModellingCompetition. The method uses competitive associative net called CAN2 embedding piecewise linear approximation scheme. With the scheme, we can naturally estimate piecewise error distribution or heteroscedastic error distribution which may be caused by the noise involved in environmental data. For improving the CAN2 with an efficient batch learning method for reducing empirical (training) error, we introduce ensemble method for more accurate prediction or less prediction (generalization) error. We also introduce multiple K-fold cross-validation for obtaining reliable predictive distribution.

show abstract

Batch learning competitive associative net and its application to time series prediction

Cited by 11 publications

References 9 publications

Improving generalization performance of bagging ensemble via Bayesian approach

Improving generalization performance of bagging ensemble via Bayesian approach

Forecasting Using First-Order Difference of Time Series and Bagging of Competitive Associative Nets

Ensemble of Competitive Associative Nets and Multiple K-fold Cross-Validation for Estimating Predictive Uncertainty in Environmental Modelling

Contact Info

Product

Resources

About