Raag detection in music using supervised machine learning approach
Ekta Patel and Savita Chauhan
Abstract
There are several research work is in progress in the direction of Raag detection. Raag is one of the melodic modes used in traditional South Asian music genres such as Indian classical music and qawwali. It can be said that Indian classical music is always set in a rāga. Non-classical music such as popular Indian film songs and ghazals sometimes use raagas in their compositions. There are several obstacles in accurate Raag detection technique. The major challenges are the complex parameters like pitch and mood in the music, skipping extra tones, conversion of different data attributes and Raag tempo. In this paper different classifiers like Bayesian net, naive Bayes, support vector machine (SVM), J48, decision table, random forest, multi-layer perceptron and PART performance are analyzed. The music features are extracted using MIRToolbox in MATLAB. These extracted features are arranged in .arff file format. WEKA tool is used. The results shown below clearly indicate that the accuracies of all the classifiers after the discretization have increases considerably. While the accuracy of the probability based classifier are best in this Raag detection from music.
Keyword
Raag, Thaats, Naïve bayes, Decision tree, Support vector machine (SVM).