Komparasi 3 Metode Algoritma Klasifikasi Data Mining Pada Prediksi Kenaikan Jabatan

Abstract

With the increasing growth of the company and employee access to increase the desire to make exemplary employees or become employees who have high ideas to become role models for subordinates or other employees, so from the research case on the university campus, quality conducts a survey with data obtained directly from the university. Employees who work as permanent or contract employees sometimes get the right as an increase for promotion from the company for each field as well as an allocation of satisfactory employee performance from the aspect of the work carried out. In this research model using three models from Naïve Bayes, K-Nearest Neighbor, and Neural Network by taking the dataset directly from the analysis results, for that an analysis is carried out on each aspect to determine the results of the value classification used in the evaluation using 5-Fold validation, 10-Fold, and 20-Fold Cross Validatio thus obtain results to identify in the promotion classification with the highest value of accuracy of 76.6%, the highest value of F1 of 67.8%, the highest value of precision of 65.9%, and the highest value of recall of 76.6%.