Derma Diseases Classification Menggunakan Metode Artificial Neural Network Backpropagation

Abstract

Skin is one of the important organs possessed by humans. The human body needs skin as the outermost protective layer from other body organs, so it is likely that the skin will be affected by disease more often. Current technological advances allow skin health experts to develop a system that can identify and classify skin diseases or derma diseases, but in reality, up to now, skin health experts still find it difficult to determine parameters and algorithms for the classification process. This research was carried out in order to develop an expert system or machine learning that can identify and classify types of derma diseases by going through the process of training and testing an artificial neural network system using the Matlab program. Expert system training and testing parameters use image projections from the extraction of morphological characteristics of research objects. The research process refers to the Artificial Neural Network (ANN) method with stages of data set preparation, image processing, training, and testing. The level of accuracy in this study was classified as high with accuracy results of 100% in the training stage and 83.3% in the testing stage.