Genre_classifier

This projects explores the use of machine learning methods such as K-means Clustering and K-Nearest Neighbors (K-NN) to predict an artist genre using audio features extracted from Spotify.

Calling Spotify API for Song Dataset, Finding Genres

1. Genre_Finder.py
2. Genre_Finder.ipynb
3. Genre_Finder.html

Cleaning Dataset, K-Means Clustering and K-Nearest Neighbors Model

1. Cleaning_and_ML.py
2. Cleaning_and_ML.ipynb
3. Cleaning_and_ML.html

Used the stats.shapiro function to test for assumptions of normal distributions in each song metric

Evaluating the best value of k for prediction accuracy

Plotting the Misclassification Error against the number of k neighbors

Confusion Matrix

Precision and recall values for each label

Total precision and recall