Customer churn can take different forms, such as switching to a competitor's service, reducing the number of services used, or switching to a lower cost service. This customer churn model enables you to predict the customers that will churn.
Many telecommunication companies track churn as part of the annual reports. In this experiment, you will learn how to create a customer churn model using sample data from the [KDD 2009 Cup](http://www.sigkdd.org/kdd-cup-2009-customer-relationship-prediction). Specifically, you will learn how to use classification algorithms (Two-Class Boosted Decision Tree, Two-Class Decision Forest) to build a model that will predict whether a customer will churn. **Inputs** The input data consists of 50,000 rows. Each row has 230 columns. The first 190 columns contain numerical data and the remaining 40 columns contain categorical data. **Outputs** The output are two columns which indicates whether a customer will churn, and the scoring probability. More details on the model are available in the book: [Predictive Analytics with Microsoft Azure Machine Learning: Build and Deploy Actionable Solutions in Minutes](http://www.amazon.com/dp/1484204468) **Related Resources** - [Power BI dashboard](https://powerbi.microsoft.com/en-us/industries/telco/) - Analyzing Telco Customer Churn. - [Video](https://azure.microsoft.com/en-us/documentation/videos/harness-predictive-customer-churn-models-with-cortana-analytics-suite/) to learn how to build the end to end customer churn solution using Cortana Analytics Suite - [Video](http://channel9.msdn.com/Shows/Azure-Friday/Azure-Data-Factory-102-Analyzing-complex-Churn-Models-with-Azure-Data-Factory) to gain a high-level overview of how analytics services in Azure can be used in a customer churn solution. Created by a Microsoft Employee