Menentukan Parameter K pada K-Nearest Neighbors (KNN) Menggunakan Random Grid Search

Abstract

This research aims to optimise the classification of trainee acceptance using the K-Nearest Neighbors (KNN) algorithm by finding the optimal value for the K parameter through a combination of Random Search and Grid Search methods. This approach is expected to overcome the problem of humanity in the trainee selection process which is subject to human error. The grid search method is used to find the parameter value sequentially, while the random search method is used for the efficiency of finding the optimal value randomly. The results show that random grid search obtained 91% accuracy with K = 7, using Manhattan metric, search time 0.015 seconds and MAE 0.085. These findings contribute to improving the time efficiency and process of classification by automatically accepting trainees using a machine learning approach, with results relevant to future system development.