JINGLECRAFT - MACHINE LEARNING IN MUSIC (2022)
- Conducted comprehensive analysis of machine learning algorithms and data visualization techniques for music genre classification and mood prediction.
- Leveraged multiple data sources including the Million Song Dataset, GTZAN, and Spotify API to create a robust training dataset. The analysis incorporated both spectral features (mel-spectrogram visualization shown below) and metadata features to capture comprehensive musical characteristics.
- Implemented various machine learning algorithms to predict song mood and emotional characteristics. The visualizations below demonstrate key insights derived from the analysis, showcasing the relationship between musical features and emotional content.
 |
 |
- This project was completed as part of CS 7641 (Machine Learning). The comprehensive analysis and findings are documented in the complete report.