Music-Cluster-Analysis

🎧 Mood-Based Music Analysis & Playlist Generator

This project analyzes songs based on their musical features (like energy, valence, and acousticness), clusters them into mood categories, and creates curated playlists for each mood. It includes an interactive Streamlit web app for exploring the playlists and mood profiles.

📌 Features

• 🔍 Data Preprocessing from a Kaggle music dataset

• 🧠 Mood Clustering using KMeans on energy, valence, danceability, and more

• 📊 Visualizations:

  • Radar charts for cluster mood profiles

  • Bar plots for top songs by acousticness, energy, etc.

• 💽 Top Track Rankings within each mood cluster

• 🌐 Interactive Streamlit Web App:

  • Explore songs by mood

  • Download mood-based playlists

  • 🔗 Directly create a Spotify playlist from your favorite mood!

🚀 Tech Stack

• Python

• Pandas for data manipulation

• Scikit-learn for KMeans clustering

• Matplotlib / Seaborn for visualization

• Streamlit for web app interface

• Dotenv for secure environment variable management

🎯 Mood Clusters Used

Each song was assigned to one of the following clusters based on its features:

1. 'Feel-good Hits' | High energy, high valence — perfect for upbeat moods |

2. 'Chill & Acoustic' | Low energy, high acousticness — calm and emotional |

3. 'Angry / Intense' | High energy, low valence — aggressive and fast |

4. 'Balanced / Versatile Mix' | Medium values across features — versatile listening |