https://github.com/alpha597/music_classification_ml

A project which compares different machine learning algorithms' accuracy in music genre classification of a large dataset.
https://github.com/alpha597/music_classification_ml

machine-learning pandas python scikit-learn tensorflow

Last synced: 2 months ago
JSON representation

A project which compares different machine learning algorithms' accuracy in music genre classification of a large dataset.

• Host: GitHub
• URL: https://github.com/alpha597/music_classification_ml
• Owner: alpha597
• Created: 2025-02-15T16:09:08.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-02-15T16:56:48.000Z (over 1 year ago)
• Last Synced: 2025-02-15T17:30:29.416Z (over 1 year ago)
• Topics: machine-learning, pandas, python, scikit-learn, tensorflow
• Language: Jupyter Notebook
• Homepage:
• Size: 166 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Music_Classification_ML

Classify music dataset using Machine Learning algorithms using Tensorflow and Skicit-learn

this classification is done by several features of music- tempo, loudness, beats, pitch, root mean square error, spectral_centroid, zero_crossing rate etc.

Extra features extracted from audio files, using librosa library of Python 

Classifying music into 10 generes - Classical , blues, rock, country, metal , pop, jazz etc. Each music genre has it's own feature combination and we  want to classify the dataset into 10 groups. At first we have used the Support Vector machine(SVM) algorithm both linear and rbf model

Collected music dataset from  https://www.kaggle.com/datasets/insiyeah/musicfeatures/data

below is accuracy of this model -confusion matrix

SVM model accuracy - approx 60% 

![image](https://github.com/user-attachments/assets/562755c1-d755-471f-aef3-b078226e4608)

next algorithm to apply - KNN, K-Means Clustering, Neural networks... etc

SVM performance can be increased using unsupervised learning PCA algorithm 

Next applied algorithm is KNN(K Nearest neighbours)

Here accuracy increases to 65%

below the new confusion maxtrix

![image](https://github.com/user-attachments/assets/493c674b-68f5-4179-8e6f-7305271d5b7b)

Next used model is KMeans Clustering, accuracy increases much more

using sklearn evaluation matrix

![image](https://github.com/user-attachments/assets/86d232a1-f59f-46a1-85b6-0d8e8f248db5)