Dielektrika : Jurnal Ilmiah Kajian Teori dan Aplikasi Teknik Elektro
Vol 8 No 2 (2021): DIELEKTRIKA

MULTICLASS CLASSIFICATION OF SOUND HEALING WITH K-NEAREST NEIGHBOR ALGORITHM

Cipta Ramadhani (Universitas Mataram)
I Made Budi Suksmadana (Unknown)
Suthami Ariessaputra (Unknown)
I Gede Pasek Suta Wijaya (Unknown)



Article Info

Publish Date
30 Aug 2021

Abstract

Sound healing can be described to the practice of sound vibrations in an individual body directly to bring about a state of harmony and healing. In many ancient country, sound healing was used as a part of medicine and healing ritual. In this paper, we propose K-Nearest Neighbors (KNN) method to categorize the type of sound healing. Acoustic Sound for Wellbeing (ASW) such as Drums, Gongs, Chimes and Singing Bowls are used as dataset for KNN algorithm. The KNN algorithm is applied to classify The ASW dataset in multi class classification tasks. In our model, KNN gave the best performance measurement for 2 Classes classification. the value of Accuracy, Precision and recall are higher than 0,87. Meanwhile, The confusion matrix for 3 classes presented the lowest point from all experimental setting. Furthermore, confusion matrix for 4 classes showed some anomaly.

Copyrights © 2021






Journal Info

Abbrev

dielektrika

Publisher

Subject

Computer Science & IT Electrical & Electronics Engineering Energy

Description

The Aims and scope of the Dielektrika are Power System, Telecommunication, electronics and computer of informatics, including: Electrical Power Systems, High Voltage Technology, Renewable Energy, Power Electronics, Sensing and Automation, Telecommunication system and technique, Signal Processing, ...