Open AccessPerformance Analysis of the RBF-SOM Network for Iris Data Classification as an Effort to Overcome System Control Problems
Author(s) -
Yoan Elviralita,
Asrul Hidayat
Publication year - 2022
Publication title -
motivection
Language(s) - English
Resource type - Journals
eISSN - 2685-2098
pISSN - 2655-7215
DOI - 10.46574/motivection.v4i1.104
Subject(s) - computer science , artificial neural network , artificial intelligence , pattern recognition (psychology) , radial basis function , matlab , self organizing map , operating system
One way to solve system control problems is by using pattern recognition. Many studies are related to pattern recognition, including artificial neural networks. This study develops an algorithm that combines artificial neural networks with Radial Basis Function (RBF) and Self-Organizing Maps (SOM). The proposed RBF-SOM algorithm was successfully realized with the MATLAB routine program and tested with the case of iris data recognition. The results of the recognition rate show that the developed artificial neural network has a good performance with an average of 98%.
Salah satu upaya dalam menyelesaikan permasalahan pengendalian system adalah dengan melakukan pengenalan pola. Banyak penelitian yang terkait dengan pengenalan pola diantaranya dengan jaringan syaraf tiruan. Penelitian ini mengembangkan sebuah algoritma perpaduan antara jaringan saraf tiruan Radial Basis Function (RBF) dan Self-Organizing Maps (SOM). Algoritma RBF-SOM ini berhasil direalisasikan dengan program MATLAB dan diuji dengan kasus pengenalan data bunga iris. Hasil recognition rate menunjukkan bahwa jaringan saraf tiruan yang dikembangkan tersebut memiliki performa yang baik dengan rata-rata sebesar 98 %.