PCA-RBPNN UNTUK KLASIFIKASI DATA MULTIVARIAT DENGAN ORTHOGONAL LEAST SQUARE (OLS)

Oni Soesanto

Abstract


This study will examine the PCA-RBPNN (Principal Component Analysis-Radial Basis) Probabilistic Neural Network) for the classification of multivariate data. The Main Component Analysis (PCA) has widely known in statistics as a method used to reduce the input dimension of the data multivariate by minimizing information loss. In this case, PCA is used to reduce dimensional input on the RBPNN neural network. The clustering process and initialization center is done with Self-Organizing Map (SOM). For the determination of weights during the learning process on the RBPNN network, using the Orthogonal Least Square (OLS) algorithm. Furthermore, PCA-RBPNN method is used for the classification of multivariate data. Accuracy of PCA-RBPNN classification is simulated and compared with the usual RBPNN model.

Keywords


RBPNN, PCA-RBPNN, SOM, Orthogonal Least Square (OLS)

Full Text:

PDF

References


Diamantaras, K.I. dan Papadimitrious, T. 2009. Applying PCA neural models for

blind separation signals’, Neurocomputing 73, 3–9.

Hasanuddin dan Irawan, I.M. 2009. The study of Sensitivity of Radial Basis

Probabilistic Neural Network, International Conference on Mathematics,

Statistics and their Application, 344-349

Hynar, M., Burda, M. dan Sarmanova, J. 2005. Unsupervisied Clustering with

Growing Self-Orginizing Neural Network - A Comparation with Non-Neural

Approach, Computer Science, 58–68.

Li, Z.X. dan Guang, H.G. 2007. Forecasting Approach for Short-term Traffic

Flow based on Principal Component Analysis and Combined Neural Network,

System Engineering 27, 167–171.

Lu, W.Z., Wang, W.J., Wang, X.K., Yan, S.Y. dan Lam, J.C. 2004. Potensial

Assesment of A Neural Network Model with PCA/RBF Approach Forecasting

Pollutant Trends in Mong Kok Urban Air, Hong Kong’, Environmental Research

, 79–87.

Huang, D.S. dan Zao, W.B. 2005. Determining the Center of radial Basis

Probabilistic Neural Network by Recursive Orthogonal Least Square Algorithms,

Applied Mathematics and Computation 162, 461–473.

Huang, D.S. dan Du, J.X. 2008. A Constructive Structure Optimization

Methodology for Radial Basis Probabilistic neural Network, IEEE Transactions

on Neural Network 19, 2009-2115.

Mu, T. dan Nandi, A.K. 2007. Breast Cancer Detection from FNA using SVM

with Different Parameter Tuning System and SOM-RBF Classifier, Journal of

Franklin Institute 344, 258-311.

Ranaweera, D.K., Hubele, N.e. dan Papalexopoulos, A.D. 1995. Application of

Radial Basis Function Neural Network Model for Short-Term Load Forecasting,

IEE Proc-Gener. Transm. Distrib. 142, 45-50.

Jurnal Matematika Murni dan Terapan Vol. 4 No.2 Desember 2010: 51 - 60

Thomassey, S. dan Happiette, M. 2007. A Neural Clustering and Classification

System for Sales Forecasting of New Apparel Item, Applied Soft Computing 7,

–1187.

Wang, X.Z., Li, C.G., Yeung, D.S., Song, SJ dan Feng, H.M. 2008. A Definition

of Partial Derivative of Random Functions and Its Application to RBFNN

Sensitivit Anaysis, Neurocomputing, 71, 1551–1526.

Yousef, R. dan Hindi, K. 2005. Locating Center Points for Radial Basis Function

Network Using Instance Reduction Techniques, Engineering and Technology 4,

-216.




DOI: https://doi.org/10.20527/epsilon.v4i2.64

Refbacks

  • There are currently no refbacks.


Copyright (c) 2017 JURNAL MATEMATIKA MURNI DAN TERAPAN EPSILON

Indexed by:

          

 

EDITORIAL OFFICE 

           

 

 

 

Creative Commons License
JMMTE is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.