predicting_customer_churn_12GB_AWS_EMR

Sparkify is an imaginary music app company and this dataset is provided as part of Udacity course. We do analysis on a small dataset followed by a larger dataset (12GB).

We analyse this datset and train a model to predict customers that are likely to churn. We see their last sessions, thumbs up, thumbs down, avg session timings and many other features to train model.

Tools Used:

1. PySpark
2. AWS EMR

Steps :

1. We first do analysis on small subset of data provided.
2. We then uses AWS EMR for 12 GB dataset.
3. We used 4 cluster and uploaded our script file on S3.