Machine learning approach for Acute Respiratory Infections (ISPA) prediction: Case study Indonesia

Abstract

The progress of human civilization is growing rapidly in all areas of life. Science and technology are hereby inseparable from the life of modern society and capital. Can't be denied the results of the life of modern society and capital spawned industries that have a big impact to the cause of the Acute Respiratory Infections (ISPA). Acute Respiratory Infections (ISPA) is a disease that attacks the breath through the nose, throat including adnexa network like sinuses with symptoms such as fever, sore throat and cough. In Indonesia ISPA ranks first cause of death in infant group and toddlers with percentages 38, 80% of all toddler mortality. ISPA is a major cause of patient visits in health services, 40-60% visit at community health clinic and 15-30% visit treatment at hospital inpatient or outpatient. In this research, we collect the dataset we get from Community Health Clinic around Denpasar and Tabanan, Bali, Indonesia. The dataset consists of 14 attribute and 150 rows of data. In this study we conducted a study to predict and diagnose disease in a person whether positive suffering from ISPA or negative in accordance with the symptoms caused by using the machine approach. In this research we will make comparisons from some algorithm is KNN, SVM, Naïve Bayes and with Neural Network. The result this research the best performance is SVM with value sensitivity 100%, specificity 100% and accuracy 100%.