Customer Churn Prediction Using Machine Learning

Customer Churn Prediction using machine learning

If you are interested in applying to GGI's Impact Fellowship program, you can access our application link here.

---

Project Objective & Business Context

Problem: Telecom companies face revenue loss due to customer churn. Goal: Predict customers likely to churn so the company can take proactive retention steps.

Use Case: Supports churn-reduction campaigns through targeted interventions.

Project Workflow Overview:

• Data Understanding (EDA)

• Preprocessing

• Train-Test Split

• SMOTE (Handling Class Imbalance)

• Model Building (Logistic Regression, Random Forest)

• Cross-Validation

• Hyperparameter Tuning

• Model Evaluation & Saving

• Model Benchmarking with LazyPredict

• Prediction on Unseen

1.Dataset Insights(EDA)

2.Feature Patterns vs Churn

3.Correlation Heatmap

4.Data Preprocessing

5.Train-Test Split: Splitting into T2 (Train) and T3 (Test)

6.Logistic Regression Results

7.Random Forest Results

8.Model Stability 3 5-Fold Cross Validation

9.Finding the Best Settings (Hyperparameter Tuning)

10.Model Deployment

11.Final Model Tested on New Unseen Data

Meet The Thought Leader

Neeraj Rajkumar Parmaar is an accomplished Artificial Intelligence Engineer currently working at KLA in Chennai, Tamil Nadu. With over 3 years of progressive experience in AI and algorithm development, he has built a strong foundation in machine learning, data science, and biomedical imaging applications.

Neeraj holds a Master's degree in Bioinformatics and Integrative Data Science from the prestigious IIT Madras, where he specialized in the intersection of computer science and biological data analysis. His academic journey also includes international exposure through studies at Aalto University in Finland.

Meet The Authors (Feynman Fellows)

Kiran Yadav is a banking operations strategist with over nine years of experience in managing a ₹100 Cr retail loan portfolio, streamlining credit risk processes, and driving cross-functional initiatives for operational excellence. She has delivered ₹40 Cr portfolio growth through data-led cross-selling and risk management, while maintaining 99% on-time, RBI-compliant disbursals. Her work spans process mapping, KPI-driven decision-making, and stakeholder alignment to deliver measurable business impact. She recently developed a machine learning model to predict customer churn, enabling proactive retention strategies.

If you are interested in applying to GGI's Impact Fellowship program, you can access our application link here.

---