Telecom-Churn-Prediction-System-CHURNLYTICAL

CHURNLYTICAL predicts customer churn output in a telecom company.

Table of Contents

Introduction

Prerequisites and Techstack

Steps for execution

How to Use the files

Usage

Sample Screenshots

Introduction

• The project CHURNLYTICAL predicts customer churn output in a telecom company.

• To implement this machine learning techniques, such as decision tree classifier, catboost classifier, etc, are used.

• To build CHURNLYTICAL prediction model two main parts were implemented:

  • Exploratory Data Analysis (EDA)
  • Model Building
  
  

• Project Structure

  The project is structured as follows:

  • churnlytical.py: This is the main application file where Streamlit code is implemented to run the application.
  • Telco_churn_Analysis_EDA.ipynb: This Jupyter Notebook contains Exploratory Data Analysis (EDA) of the telecom churn dataset.
  • Telco_churn_Analysis_Model_Building.ipynb: This Jupyter Notebook contains the code for building and training the churn prediction models.

back to top

Prerequisites and Techstack

• Python (version 3.8)
• Streamlit (version 1.14.0)
• Pandas (version 1.4.3)
• Scikit-learn (version 1.0.2)

back to top

Steps for execution

To run the application, execute the following command in the terminal:

streamlit run churnlytical.py

This command will start the Streamlit application and open it in your default web browser(localhost:8501).

back to top

How to Use the files

1. EDA: To explore the Exploratory Data Analysis (EDA) of the telecom churn dataset, refer to the Telco_churn_Analysis_EDA.ipynb notebook.

2. Model Building: To build and train the churn prediction models, refer to the Telco_churn_Analysis_Model_Building.ipynb notebook.

3. Prediction:

   • To predict churn for a single customer, use the Streamlit application (churnlytical.py). Input the customer's information, and the application will provide the churn prediction.
   • To predict churn for a batch of customers, prepare a CSV file in the format of the sample.csv provided. The CSV file should contain the customer information. Then, use the Streamlit application (churnlytical.py) to upload the CSV file and get the churn predictions for the batch.

4. Dataset:

   The dataset used for this project can be found here, and it contains the necessary information for training and testing the churn prediction models.

back to top

Usage

• This project can predict customer churn in a telecom company using machine learning techniques. It includes two main parts: Exploratory Data Analysis (EDA) and Model Building.
• The project is designed to facilitate prediction for a single customer or a batch of customers by providing a CSV file for the batch.

back to top

Sample Screenshots

• Home page

  
  

• Login-Logout page

  
  

• Churn Prediction for One customer page

  
  

• Churn Prediction for Many customers page

back to top

Thank you for exploring the CHURNLYTICAL project.