Influence Optimization Feature Against Liver Disorders Diagnostic Results Using Artificial Neural Network

Abstract

The diagnosis is a classification of a person based on a disease or abnormality . One classification technique that can be used is Artificial Neural Networks. ANN is an information processing system that has characteristics similar to human nerves, in ANN training data is needed in learning. The learning process in Artificial Neural Networks related to the length of time the learning is done. One way to reduce computing time can be done with the selection feature. In this study, an analysis of the results of the diagnosis of liver disorders using Artificial Neural Networks with feature selection and without feature selection. The test results show that the accuracy of the data obtained by performing feature selection tends to be more stable when compared to the value of data accuracy without feature selection. Besides the learning time required by the data that do feature selection tends to be faster than data that does not do the feature selection.