Abstract

Sound healing, or commonly referred to as music therapy using Acoustic Sound for Wellbeing (ASW) equipment such as drums, gongs, bells, and other types that produce specific frequency vibrations, is used in the medical field to help patients experiencing anxiety or depression. Currently, research on sound healing focuses on methods to identify appropriate frequencies that influence stress and anxiety experienced by patients. This study presents the implementation of the Pseudo-Nearest Neighbour (P-NN) algorithm for classifying multiclass ASW. In general, the P-NN algorithm performs better for multiclass scenarios, particularly in identifying outlier data in each class. Furthermore, P-NN provides better performance for all confusion matrix parameters. Using two classes (Gong and Singing Bowl), the accuracy of the P-NN algorithm exceeds 92%. This demonstrates that the P-NN algorithm can provide better performance in handling outliers within the ASW dataset.

Concepts :

Access to Document

Citations by Year