IMPLEMENTASI METODE SELEKSI FITUR MENGGUNAKAN ARTIFICIAL BEE COLONY PADA KLASIFIKASI RETINAL NERVE FIBER LAYER

Abstract

Damage to Retinal Nerve Fiber Layer can Cause Glaucoma. Glaucoma is an inflammation of the optic eye which is characterized by progressive deterioration of Optic Nerve Head and field of view. Problems that require a classification solution are hindered by the large data dimensions. Artificial Bee Colony is one of the evolution algorithms widely used for feature selection and optimization. Gray level Coocurrence matrix is used as a feature extraction method, the Artificial bee colony method is used as a feature selection and Support Vector Machine used as Classification. The proposed method using Artificial Bee Colony gets improved Accuracy compared to method without using Artificial Bee Colony. The results obtained by the proposed method were 95% for accuracy, 95.9% for specificity and 93.7% for sensitivity where methods that did not use Artificial Bee Colony obtained an accuracy of 93.8%, Specificity sebsar 90.3% and Sensitivity of 92.6%.

Keywords

• RNFL, GLCM, Artificial Bee Colony, Feature Selction, Classification