Data Mining untuk Klasifikasi Status Gizi Desa di Kabupaten Malaka Menggunakan Metode K-nearest Neighbor

Classification of village status according to the number of malnourished patients is very important in anticipating malnutrition cases in a region, especially for the areas in the district of Malaka. Cases of malnutrition recorded quite a lot in the District of Malaka demanded the district government of Malaka to immediately anticipate the problem. To overcome this problem, we used k-Nearest Neighbor method to classify the status of villages in Malaka District based on the level of under-five children under the red line into three target classes: low, medium, and high. Prior to the classification process, clustering process is done using K-Means method so that all data can be divided into classes that have been determined. The data used in this study as many as 174 data taken from the year 2013-2015. The final result, after validation of clustering data obtained resemblance to the original data of 98.25%, and the results of system testing of 93.10%. Determination of the best value of k with the test data of 34 pieces and the training data of 140 pieces is at k = 7 with the average percentage of similarity of 95.53%