Desain dan Analisa Pengklasifikasian Data Wajah Menggunakan Metode Incremental Radial Basis Function dengan Modifikasi Unit Hidden

Abstrak

This study implemented the improvement of the network learning method, that is, the development of the Incremental Radial Basis Function (IRBF) network with the data used in this study was facial data. The classification of data on facial data was strongly influenced by the characteristics of the data, so that it affected the success of data recognition. This study showed that the percentage of learning success using the Radial Basis Function method, the learning data succeeded in identifying the classification of facial data with an average trial result of 99.45%. The identification of facial data learning with Incremental Radial Basis Function Development was 99.45% on learning data. In addition, the data for testing trials gained an increase in data with the average test results from 89.44% in the RBF method to 90.55% in the Modified Gaussian Incremental Radial Basis Function (MGIRBF) method. The method of developing the Incremental Radial Basis Function with a modified gaussian in this research was applied as a development after applying the Radial Basis Function method by prioritizing optimization by applying the characteristics of the Incremental Radial Basis Function method.