Open AccessRancang Bangun Sistem Pakar untuk Deteksi Dini Katarak Menggunakan Algoritma C4.5
Author(s) -
Ivana Herliana Winarsih Jayawardanu,
Seng Hansun
Publication year - 2016
Publication title -
ultima computing/ultima computing: jurnal sistem komputer
Language(s) - English
Resource type - Journals
eISSN - 2549-4007
pISSN - 2355-3286
DOI - 10.31937/sk.v7i2.232
Subject(s) - blindness , indonesian , training set , computer science , test set , set (abstract data type) , artificial intelligence , medicine , test (biology) , optometry , algorithm , machine learning , philosophy , linguistics , programming language , paleontology , biology
In 2010, 51% of 39 million blindness are caused by cataract. In 2013, there are 1.8% of 1.027.763 Indonesian people who suffered from cataract. Half of them are not treated yet due to their ignorance on the cataract disease. Therefore, in this research, we tried to build a system that can detect early cataract disease as the ophthalmologist would do. The system will use C4.5 algorithm that receives 150 training data set as an input, resulting in a set of rules which can be used as decision factors. To test the system, k-fold cross validation technique is been used with k equals to 10. From the analysis result, the accuracy of the system is 93.2% to detect cataract disease and 80.5% to detect the type of cataract disease one might suffered.
Index terms-C4.5 algorithm, cataract, k-fold cross validation, machine learning