Open AccessMachine Learning Algorithms for Indian Music Classification Based on Raga Framework
Author(s) -
Kalyani Waghmare,
Balwant A. Sonkamble
Publication year - 2020
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.k7724.0991120
Subject(s) - artificial intelligence , computer science , machine learning , unsupervised learning , cluster analysis , decision tree , support vector machine , naive bayes classifier , pattern recognition (psychology)
The supervised and unsupervised learning methods in Machine Learning are successfully applied to solve various real time problems in different domains. The Indian Music has a base of Raga structure. The Raga is melodious framework for composition and improvisation. The identification and indexing of Raga for Indian Music data will improve efficiency and accuracy of retrieval being expected by e-learners, composers and classical music listeners. The identification of Raga in Indian Music is very difficult task for naïve user. The application of machine learning algorithms will definitely be best key idea. The paper demonstrates K-means and Agglomerative clustering methods from unsupervised learning nonetheless K Nearest Neighbor, Decision Tree and Support Vector Machine and Naïve Bayes classifiers are implemented from supervised learning. The partition of 70:30 is done for training data and testing data. Pitch Class Distribution features are extracted by identifying Pitch for every frame in an audio signal using Autocorrelation method. The comparison of above algorithms is done and observed supervised learning methods outperformed.