PENELITIAN AWAL PENGELOMPOKKAN DATA PENDUDUK MEMPERGUNAKAN ALGORITMA K-MEANS CLUSTERING UNTUK PENENTUAN PENERIMAAN BANTUAN DANA DESA

Abstract

The problem of determining who is entitled to receive village fund assistance by a village government often has an impact on the unsuccessful purpose of providing funds, because it can result in families who should have received village fund assistance becoming unable to obtain their rights.  One of the things that can be taken is to utilize the K-Means clustering algorithm to assist decision makers in choosing residents who are eligible to receive the funding assistance, in the hope that the results can be more accurate. From the results of the experiment using 188 population data in village X who had passed the initial data processing including removing irrelevant attributes, the largest silhouette value in the results of grouping data into 2 parts was 0.728278964. The value indicates that the results of the data grouping that occurred were already strong and each data was already placed in a very suitable cluster. The attributes that are thought to be used as a basis for determining decision making are the number of families, land area, floor area, type of floor, wall material, presence of windows, roofing material, type of lighting, garbage disposal, toilet facilities, water sources, and disposal facilities. From the ground truth value found at 70.21%, it can also be seen that the results of the grouping are relevant to reality in the field.