Using machine learning to reduce customer churn
This tutorial demonstrates a decision service that recommends retention offers based on
predicted customer churn and lifetime value. These predictions are generated by two predictive
models, each connected to Watson Machine Learning deployments created with AutoAI.
Learning objectives
You will learn how to:
- Build a Watson Machine Learning model and deploy it using an AutoAI experiment.
- Configure a Watson Machine Learning provider in Decision Designer and connect it to a predictive model.
- Run a decision model that uses this predictive model.
Audience
This sample is designed for both technical and business users who want to leverage predictive analytics through machine learning in decision services using Decision Intelligence Client Managed Software. It also demonstrates to data scientists and data engineers how Decision Intelligence can integrate machine learning models into decision-making applications.
Time required
15 minutes
Prerequisites
Before you begin, make sure you review the following resources:
- Quick guide: Integrating Watson Machine Learning with decision services
- Deploying AI assets using Watson Machine Learning
You will also need access to the following environments:
- Decision Designer: A web-based interface for developing decision services.
- Watson Studio: A web-based platform for creating and deploying machine learning models.
Download the customer_churn_data.csv and customer_LTV_data.csv data sets to build machine learning models using
AutoAI in IBM Watson Studio.