Seminar Nasional Aplikasi Teknologi Informasi (SNATI)
2005

Pengenalan Huruf Menggunakan Model Jaringan Saraf Tiruan Radial Basis Function dengan Randomize Cluster Decision

Muhammad Erwin Ashari Haryono (Unknown)



Article Info

Publish Date
02 Oct 2009

Abstract

Artificial Neural Network is the adaptive model in Artificial Intelligence as become the model of learning.Many research in artificial neural network using BackPropagation. More of 70 prosen researcher using thismodel to solve all problem in its research. BP using supervised learning, which indicated the output as the basisof failure value as a feedback of its activation. This paper I will introduced another method in artificial neuralnetwork called Radial Basis Function to recognize various font. RBF is a mocel of ANN which adopt bothsupervised and unsupervised learning. The Hidden layer doesn’t form first when the building of its architecture.In this model Hidden layer will form after we could cluster for all various font itself. A number of hidden layerbased on a number of cluster center. Weighting in Hidden and output layer using Gaussian model andPseudoinverse matrices. Although I didn’t compare the result of both model BP and RBF for this problem, theresult of this research almost get 70 % success in order to recognize all font pattern.Keywords: Artificial Neural Network, adaptive, Radial Basis Function, Gaussian, Pseudoinve

Copyrights © 2005