Logo Lanfrica

elvis-darko/Team_Zurich_Capstone_Project

Domain:

socioeconomic

Record type:

project
Creator:
elv
Host:
This repository contains the final project of the Azubi Africa Data Science Apprenticeship Training Programme. # EXPRESSO CUSTOMER CHURN PREDICTION ## PROJECT DESCRIPTION This project is for an African telecommunications company that provides customers with airtime and mobile data bundles. The objective of this project is to develop a machine learning model to predict the likelihood of each customer “churning,” i.e. becoming inactive and not making any transactions for 90 days. This solution will help this telecom company to better serve their customers by understanding which customers are at risk of leaving. ## SUMMARY | Code | Name | Deployed App | |-----------|-------------|:------| |CP | EXPRESSO CUSTOMER CHURN PREDICTION | STREAMLIT APP| # NOTES ON DATASETS The complete dataset can be found and downloaded from Zindi: Zindi Churn Challenge **Train data :** Contains information about 1 million customers. There is a column called CHURN that indicates if a client churned or not. This is the target variable. You must estimate the likelihood that these clients churned. We will use this file to train our model. **Test data :** This is similar to the train data but without the Churn column. We will use this file to test our model. The dataset has 19 variables. Below are the definitions of variables found in the datasets: |VARAIABLE|DEFINITION|FEATURE| |---------|----------|---------| |`user_id`| A unique identification number of a client|Numeric| |`REGION` |The location of each client|Categorical| |`TENURE` |Duration of the network usage|Numeric| |`MONTANT` |The top-up amount|Numeric| |`FREQUENCE_RECH`| The number of times the customer refilled|Numeric| |`REVENUE`|The monthly income of each client|Numeric| |`ARPU_SEGMENT`|Income over 90 days / 3|Numeric| |`FREQUENCE`| The number of times the client has made an income|Numeric| |`DATA_VOLUME`| The number of connections|Numeric| |`ON_NET`|Inter Expresso calls|Numeric| |`ORANGE`|calls to Orange network|Numeric| |`TIGO`|calls to Tigo network|Numeric| |`ZONE1`| calls to zones1|Numeric| |`ZONE2`|call …