Raga Yakshagana: Machine Learning for Yakshagana Bhagavatige's Raga Recognition

The study titled “Yakshagana Bhagavatige Raga Recognition Using Machine Learning” focuses on developing an innovative system capable of identifying distinct ragas within Yakshagana Bhagavatige performances based on audio input. Employing advanced machine learning techniques, specifically the Random Forest classifier, XGBoost, Gradient Boosting, and Logistic Regression, we meticulously train our model using a comprehensive dataset of relevant features. The primary focus of this paper is to bridge the gap between traditional Yakshagana artistry and contemporary technology. The outcome is a model that, upon receiving an audio file as input, accurately predicts the corresponding raga within the Yakshagana Bhagavatige, thus contributing significantly to the intersection of cultural heritage and machine learning. This paper details our methodology, dataset creation, training process, and the promising results achieved through the implementation of diverse classifiers, showcasing the potential for technology to enhance the understanding and preservation of cultural art forms.