PremKumar-V/Telecommunication_Churn

Prediction Project on Telecommunication Company Churn

Telecommunication_Churn

Prediction Project on Telecommunication Company Churn

References

• Dataset Link
• Sklearn Docs

Files

• Original Dataset File
• Preprocessed Dataset File
• Saved Version of Random Forest Model

KeyPoints

• Dataset contain 20 columns, where 3 Numerical and 17 Categorical.
  • On Categorical 6 Binary Values, 9 Non Binary Values and 1 Object Value.
• Analysis
  • Gender is Appox Equal with points apart.
  • Count of Senior Citizen is low.
  • Chance of Senior Citizen having phone service is low as Internet Service too.
  • Most of the Internet is Fibre Optic.
  • Not everyone has Online Backup.
• While Preprocessing Consideration of No Phone Service and No Internet Service as NO is must.
• And also Consideration of Fibre Optic and DSL as YES.
• Random Forest Algorithm scored more than Support Vector Machines