Classification of Thaats in Hindustani Classical Music using Supervised Learning

The Hindustani Classical System of music comprises various ragas, which are categorized under 10 Thaats. These thaats are classified based on the notes being majorly used in the individual ragas coming under the thaat. In these modern days of digitalization, we have various sources of information regarding these ragas, but when it comes to a seeker practicing a given raga, he may find it difficult to stay aligned with the notes that are being used, which may lead to wrong practice and eventually a mess. In this paper, we have proposed a method to classify a piece of music into one of the 10 thaats based on the notations being used in it. Every note has a corresponding frequency, analyzing which we can identify the thaat to which a given piece of music belongs. After extracting the frequencies from a piece of music, we can train our supervised learning models which can be further used to create an automated guru that in real-time will indicate whether the notes being used are correct or not.