Producing self tuned music using machine learning tools

This thesis report is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science and Engineering, 2017.Cataloged from PDF version of thesis report.Includes bibliographical references (page 20-21).Automatic music genre classification is one of the important tasks for the Music Information Retrieval (MIR). With the development of the knowledge on Machine Learning, researchers have implemented different methods to implement automatic music genre classification. In this paper, we have done secondary research on various papers that have used different mechanisms to achieve results for genre detection. We will check for the performance of genre detection using kNN (k-Nearest Neighbor Classifier) Classifier and the SVM (Support Vector Machine) classifier. We used MATLAB and specific toolboxes made by MIRLab such as Machine Learning Toolbox and Speech and Audio Processing Toolbox and used different kinds of features for classifiers such as MFCC (Mel-frequency cepstral coefficients). We will use the GTZAN dataset to do the classification using the classifiers. This goal is to see which classifier algorithm on MFCC features performs optimally on the GTZAN dataset. We have used various papers as references on this topic.Akifa TasneemMostofa Saif RezaNavid AnindyaB. Computer Science and Engineering