Cross-Validation and Validation Set Methods for Choosing K in KNN Algorithm for Healthcare Case Study

Robbi Rahim (Sekolah Tinggi Ilmu Manajemen Sukma, Jl. Sakti Lubis, Kota Medan, Sumatera Utara, 20219, Indonesia)
Ansari Saleh Ahmar (Department of Statistics, Universitas Negeri Makassar, Makassar, Indonesia)
Rahmat Hidayat (Department of Information Technology, Politeknik Negeri Padang, Limau Manis, Padang, 25164, Indonesia)

JINAV: Journal of Information and Visualization
Vol. 3 No. 1 (2022)

KNN categorization is simple and successful in healthcare. In this research's example case study, the KNN algorithm classified the new record as "Abnormal." The classification method began with choosing K, then calculating the Euclidean distance between the new record and the training set, finding the K nearest neighbors, then classifying the new record based on those K neighbors. The findings show that the KNN algorithm is effective in healthcare and highlight several shortcomings that should be addressed in future study. Weighting variables, choosing the best K value, and handling non-uniform data are these restrictions. The findings show the KNN algorithm's medical potential.