Text
PERBANDINGAN METODE RANDOM FOREST DENGAN METODE K-NEAREST NEIGHBOR UNTUK KLASIFIKASI PENYAKIT DIABETES MELITUS
Diabetes Mellitus is a disorder of the insulin system caused by excessive levels of glucose in the blood. Symptoms of diabetes mellitus so far have only been diagnosed by the general public based on known physical characteristics without being supported by facts and other medical considerations. The Basic Health Research (Riskesdas) shows a significant increase in the prevalence of diabetes, from 6.9% in 2013 to 8.5% in 2018. This study uses the Random Forest method and the K-Nearest Neighbor method where a comparison is made. to see which method is most appropriate in classifying Diabetes Mellitus. Both of these methods have the advantage that they can be used for large amounts of data. Determination of optimal parameter values is very important to support good accuracy results in both methods. Based on the results of the research, the Random Forest method produces an average accuracy of 80.91% accompanied by the average precision, recall, and f-measure values, respectively, which are 84.536%, 78.782%, and 81.422% with the best level of accuracy reaching 82.47%. While the K-Nearest Neighbor method produces an average accuracy of 74.062%, accompanied by the average value of precision, recall, and f-measure, respectively, which is 72.83%, 81.72%, and 76.992% with the best level of accuracy reaching 76.62%. This proves that the Random Forest method is the most suitable in classifying diabetes mellitus compared to the K-Nearest Neighbor method.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2207002582 | T75168 | T751682022 | Central Library (Referens) | Available but not for loan - Not for Loan |
No other version available