Analisis Kinerja Struktur Data Kd-Tree Pada Metode K-Nearest Neighbors

Yuan Lukito (Program Studi Teknik Informatika, Fakultas Teknologi Informasi)

Riau Journal of Computer Science
Vol. 2 No. 2 (2016): Riau Journal of Computer Science

K-Nearest Neighbors is a commonly used classification technique that can be categorized into instance-based classification method. The performance of KNN is mostlydetermined by the size of the training data. This research compared and analyzed KD-Tree andArray data structures on KNN implementation. Dataset used in this research has largemultidimensional features. From the experiment conducted we can conclude that KD-Tree datastructure has better and relatively stable performance compared to Array data structure.