Evaluation of different Machine Learning Algorithms for Music Genre Classification

Getting Started

This project evaluates different Supervised and Unsupervised Machine learning algorithms for classifying a song to a Genre.

Prerequisites

scipy
numpy
scikits.talkbox
pydub
sklearn

Procedure followed to prepare the train and test datasets:

Step 1: Convert the dataset in MP3 to WAV format.

Step 2: Label the train and test dataset with the respective Genre.

Step 3: Extract the features of train and test dataset using MFCC feature extraction algorithm.

Step 4: Store the extracted features as a Pickle file.

Step 5: Use various Machine Learning Algorithms to compare the Accuracy of Classification.