User-activity data records typically contain user actions performed on a website, service, or product. These user activity transactions tell us a lot about the user's current interests & preferences. Knowing these interests and preferences can help make businesses make better decisions.
Recommender systems, fraud detection, churn prediction, and lead scoring are examples of data products that can require user-activity data. In this two-part tutorial, you will first learn how to work with user activity data and then learn about two specific examples of applications that can leverage user activity data; churn prediction & lead scoring;
Churn prediction is the task of identifying users that are likely to stop using a service, product or website. Lead scoring is the task of prioritizing users based on the probability that are likely to start using a service, product or website. In the first part of the tutorial, you will learn to:
- Train a model to forecast user churn
- Explore & Evaluate predictions made by the model
- Consume predictions made by the model in an external application