Prediksi Penambahan Piutang Iuran Jaminan Sosial Ketenagakerjaan menggunakan Algoritma K-Nearest Neighbor

Abstract

There are several issues with Social Security Organizing Agency (BPJS) employment at the moment, one of which is contribution receivable. To reduce the BPJS contribution receivables, BPJS has done various ways. However, the resulting effort is not maximal enough to reduce the number of receivables in BPJS. This study aims to provide input by predicting the addition of receivables from social security contributions made by several companies or organizations. This study used the K-Nearest Neighbor (KNN) Algorithm with a cross-validation technique. KNN is a very simple classification method in classifying an image based on the closest distance to its neighbors. This study conducted data processing from BPJS use, which amounted to 1193 data. The data is then preprocessed so that the processed data is clean from missing and noise, this data uses 70:30 data splitting. After the preprocessing and splitting of data were carried out, the next step was to do modeling using KNN, so the cross-validation to improve the accuracy of results obtained from the KNN algorithm. The results obtained from this research get the highest accuracy of 92% with the Optimal K value being 6, then the ROC curve gets 94% accuracy. From these results, it can be said that the use of cross-validation can increase the accuracy of this study.

Penerapan Algoritma K-Nearest Neighbor Menggunakan Wrapper Sebagai Preprocessing untuk Penentuan Keterangan Berat Badan Manusia

“…Tahap pertama dari preprocessing adalah cleaning [14]. Data yang noise, tidak konsisten, atau tidak relevan dihilangkan melalui pembersihan data.…”