Service Area

Mining Classification Rules from a Subscriber Database for Churn Prediction

Area: Telecommunication
• Description

Churn prediction takes information about the historical behavior of subscribers and applies predictive models to determine the likelihood for churn. Make informed decisions for your company, enabling you to keep the customers you have and obtain more of the subscribers you want.

In this technology, the objective is to mine the classification rules from the subscriber database provided by a telecommunication company in order to discover interesting relationships concerning with the demographics and the behaviors of the subscribers who had churn before.

- Special Features and Advantages

  • Searches through all possible rule space to find a near-optimal rule sets with an evolutionary approach.
  • Generates high order classification rules to reveal complex inter-relationships.
  • Utilizing data mining techniques to enrich your knowledge of the subscribers.
  • Helping you understand which subscribers are candidates for churn.
  • Inherently parallel in nature and easily deployed in the cloud environment.

- Applications

  • Customer Relationship Management
    Understand subscribers’ behaviors to determine the likelihood for churn.