Churn Prediction

Customer Churn Prediction

Telco Customer Churn Prediction 👦🏻👧🏻🧑🏻👨🏻🏃🏻‍♂️🏃🏻‍♀️

• Churn Prediction is a Key Predictor of the Long Term Sucess or Failure of Business.

• Churn | Attrition : Customers Who Left using Company Product or Service within the Last Month.

• Customer Retention should be a Top Priority of any Business for keepin the Existing Loyal Customers.

• A Company should determine the Customers more at Risk and take Preventive Measures.

Data Set : Kaggle Telco Customer Churn

• Each Row Represents a Customer.

• Each Column Represents Customer’s Attributes.

1.Demographic Data :

• Age : Age of a Customer.

• Gender : Male | Female.

• Senior Citizen : 1 | 0.

• Partner or Single : Yes | No.

• Dependent or Independent : Yes | No.

2.Services of Company :

• Phone Service : Yes | No.

• Multiple Lines : Yes | No | No Phone Service.

• Intenet Service : DSL | Fibre Optics | No.

• Online Security : Yes | No | No Internet Service.

• Device Protection : Yes | No | No Internet Service.

• Tech Support : Yes | No | No Internet Service.

• TV Streaming : Yes | No | No Internet Service.

• Movies Streaming : Yes | No | No Internet Service.

3.Accounts Information :

• Contract : Month to Month | Two Year | One Year.

• Payment Method : Electronic Check | Mailed Check | Bank Transfer (Automatic) | Credit Card (Automatic).

• Paperless Billing : Yes | No.

• Monthly Charges

• Total Charges

• Tenures : Length of Tenure in Months.

4.Target :

• Churn

Packages :

1. NumPy

2. Pandas

3. Matplotlib

4. Seaborn

5. Sweetviz : Beautiful and High Density Visualizations for Exploratory Data Analysis

6. Scikit Learn :

• Preprocessing : MinMaxScalar

• Model Selection : Train Test Split

• Linear Model : Ridge Classifier

• Ensemble : Random Forest Classifier

• Metrics : Accuracy Score

• Model Improvement : Grid Search Cross Validation

Achieved an Overall Accuracy of 90%