Churn prediction assignment

Index

• Assignment Summary
• Import Libraries
• Understanding & Preparing Data
  • Data summary
  • Encoding the churn variable into 0 and 1
  • Changing TotalCharges column from object to float
  • Check for null and total observations related to it
  • summary description of the numeric variables of the dataset
  • Check the number of unique values in each of the columns
  • Calculate the proportion of churn vs non-churn
  • Churn Distribution by gender
  • Calculate the proportion of churn by contract
  • calculate the proportion of churn by payment method
  • Visualize churn by payment method
  • Proportion of churn by gender and contract
  • Observations by citizen type
  • Visualize Churn rates by Citizen Type, Tech Support & Contract Status
  • Finding correlations and plot the heat map
  • Segmenting based on data type and pre-processing
  • Creating a dataset to combine pre-processed variables
• Data Preprocessing
  • Feature importance
  • Import machine learning libraries
  • Feature extraction
• Train Test Split
• Modeling
  • Model Building
  • Model Training
  • Model Evaluation
• Summary

Assignment Summary

In this assignment I will not only look at what are the attributes for customers to terminate services, but I will also try to make an analysis for what can be done about it.