Cluster analysis on Spotify audio features
Problem definition: Examine a set of audio_features of songs and using clustering techniques identify potential genres. The goal is define genres and their characteristics.
Data file contains 1657 rows and 12 Spotify audio features:
Songs were extracted using spotify API using sportifyr R package
I used three different type of clustering:
- PAM K-Medoids clustering
- Agglomerative clustering
- Density based clustering
- PCA
Problem definition: Created segmentation on dataset of audio tracks. Extracted genres to compare results.
Comming soon!!!