Sparkify Churn Prediction

• Tech Stack: PySpark, Tensorflow, NLTK, Keras, Python, Opencv
• Github URL: Project Link

Predicted customer churn using Random Forest Classifier, Logistic Regression, GBT Classifier & Naive Bayes. Executed data transformation & feature engineering to increase the predictive power of ML algorithms by 21%. Implemented model tuning along with feature importance, which resulted in a reduction of churn rate by 35%.