Perbandingan Metode K-NN dan SVM Berdasarkan Kinerja Pegawai

Sinarring Azi Laga (Universitas Hayam Wuruk Perbanas, Surabaya)

Jurnal Sistem Komputer dan Informatika (JSON)
Vol 4, No 3 (2023): Maret 2023

Limited qualified human resources cause employees not to do the job in accordance with the company's operational standards properly and correctly. At this time PT. XYZ does not have tools to identify employee performance, therefore researchers conduct research to assist PT. XYZ in classifying employee performance. The methods used in this study were K-NN and SVM with a sample of 873 PT. XYZ employee data. Based on the trials conducted, the K-NN method has the highest accuracy rate of 90.13%, 91% precision rate, and 98.95% recall rate. The most optimal number of neighbors (k value) for the K-NN method is 5 with an accuracy rate of 88.35%.