Customer Churn Prediction Using Keras

This repository contains the analysis and model building for predicting customer churn at a telecom company named "Leo". The dataset used is 'Customer_Churn'. The main objective is to analyze the data, derive insights, and build predictive models to identify customers likely to churn

Project Notebooks

• Capstone 2 - Project.ipynb

Dataset

The dataset Customer_Churn includes various customer details, and the target variable is whether the customer has churned or not.

About

This project aims to predict customer churn for the telecom company "Leo" using Keras. It involves data manipulation, visualization, and model building to identify customers likely to churn.

Key Features

• Data Manipulation: Prepare the dataset for analysis and modeling.
• Data Visualization: Visualize customer distribution and service usage.
• Model Building: Develop and evaluate sequential models using Keras.

Tasks

A) Data Manipulation

• Find the total number of male customers.
• Find the total number of customers whose Internet Service is ‘DSL’.
• Extract specific customer segments based on criteria such as senior citizens and payment methods.

B) Data Visualization

• Build a pie-chart to show the distribution of customers who would be churning out.
• Build a bar-plot to show the distribution of ‘Internet Service’.

C) Model Building

Model 1

• Build a sequential model using Keras with tenure as the feature.
• Input layer: 12 nodes, 'Relu' activation.
• Hidden layer: 8 nodes, 'Relu' activation.
• Optimizer: 'Adam'.
• Epochs: 150.

Model 2

• Add dropout layers to Model 1 for improved robustness.

Model 3

• Use tenure, monthly charges, and total charges as features.