<![CDATA[One of the psychiatric diseases that affects many Indonesians is schizophrenia. Schizophrenia causes a person to sustain delusions, hallucinations, chaotic thoughts, and behavioral changes. According to Riset Kesehatan Dasar (Riskesdas) in 2013, prevalence of schizophrenia is 1.7% per 1000 people or about 400,000 people. For very wide territory of Indonesia with total population around 237 million, the number of psychologists or psychiatrists about 616 people is still very small. With this limitation, a system that can be used to assist paramedics in diagnosing and classifying psychiatric illnesses of schizophrenia. In this study applied fuzzy K-nearest neighbor algorithm to diagnose and classify psychiatric illness of schizophrenia. Types of schizophrenia used in this study are paranoid schizophrenia, hebephrenic schizophrenia, catatonic schizophrenia, undifferentiated schizophrenia, and simple schizophrenia. The classification process consists of three processes are the fuzzy initialization process, the K-nearest neighbor algorithm process, and the fuzzy K-nearest neighbor algorithm process. The testing consists of the effect of K value and the effect of K-Fold. Based on the test results on the K value, obtained the highest accuracy of 38.33% at K=5. The effect of K-Fold test results obtained the highest average accuracy of 34.17% at K-Fold= 10.]]>
Copyrights © 2018
