Survival Analysis

Orange Telecom’s Churn Dataset

The Orange Telecom’s Churn Dataset is designed to help in developing predictive models to identify customers who are likely to cancel their subscriptions. Here’s a detailed breakdown of the dataset:

Overview:

• Content: The dataset includes cleaned customer activity data (features) and a churn label indicating whether a customer has canceled their subscription.
• Purpose: Used for developing predictive models to anticipate customer churn.

Key Features and Columns:

1. Customer Information: Data such as customer IDs and demographic details.
2. Behavior and Usage: Details on customer behavior, usage statistics, and other relevant metrics.
3. Service Plans:
   • Voice Mail Plan (vmailplan): Indicates if the customer has a voice mail plan.
   • International Plan (intlplan): Indicates if the customer has an international call plan.
4. Usage Statistics:
   • Day Minutes (daymins): Total minutes of calls during the day.
   • Day Calls (daycalls): Total number of calls during the day.
   • Day Charges (daycharge): Charges for the calls during the day.
   • Similar metrics are available for evening, night, and international calls.
5. Customer Service Interaction: Number of calls made to customer service (custservicecalls).
6. Churn Indicator (churned): A binary indicator showing whether the customer has churned (1 if churned, 0 otherwise).

Sample Columns:

• customerID: Unique identifier for each customer.
• accountdur: Duration of the customer’s account (in months).
• areacode: Area code of the customer’s phone number.
• phonenumber: Customer’s phone number.
• vmailmessage: Number of voicemail messages.
• evemins: Total minutes of calls during the evening.
• evecalls: Total number of calls during the evening.
• evecharge: Charges for the calls during the evening.
• nightmins: Total minutes of calls during the night.
• nightcalls: Total number of calls during the night.
• nightcharge: Charges for the calls during the night.
• intlmins: Total minutes of international calls.
• intlcalls: Total number of international calls.
• intlcharge: Charges for international calls.

Example Use Cases:

• Survival Analysis: Estimating the time until a customer churns.
• Predictive Modeling: Identifying factors that contribute to churn and predicting which customers are likely to churn.
• Descriptive Statistics: Understanding the distribution of features such as service usage and customer service calls.