01
Research Design
Defined the project objectives and analytical framework for investigating customer churn.
Data Science & AICustomer Churn Prediction
BSc. Data Science Group Project · 2023
ICTlab, University of Science and Technology of Hanoi
Hanoi, Vietnam
Customer Churn Prediction
A machine-learning project exploring the factors associated with customer churn and comparing predictive models to support customer-retention strategies.
Machine Learning · Customer Analytics · Predictive Modelling · Data Mining
Project Overview
A collaborative data mining project using telecommunications data to identify customer churn patterns and compare machine-learning models for retention analysis.
Role
Project Contributor
Methodology Design
Context
Data Mining
Group Course Project
Collaboration
Hai Minh Hua
Duc Thai Phung
Xuan Trung Pham
Anh Quan Nguyen
Areas
Customer Analytics · Classification · Feature Importance
Tools & Technologies
Python · Exploratory Data Analysis · Decision Tree · Random Forest · Multilayer Perceptron
My Contribution
02
Data Analysis
Contributed to preprocessing, exploratory analysis, and interpretation of customer behaviour patterns.
03
Model Evaluation
Compared predictive models and interpreted key churn factors for retention analysis.
Dataset
Public telecommunications customer churn dataset with 7,043 customers and 21 variables
View Dataset on KaggleProject Workflow
The project moves from customer data preparation to model comparison and interpretation of churn-related factors.
Step
01
Data Preparation
Prepare telecommunications customer records for churn analysis and model training.
Step
02
Exploratory Analysis
Explore customer attributes, service usage, and churn distribution to identify analytical directions.
Step
03
Model Comparison
Compare Decision Tree, Random Forest, and Multilayer Perceptron models for churn prediction.
Step
04
Retention Insight
Interpret churn-related factors and connect model outputs to customer-retention decisions.
Project Outcome
The project identified key factors associated with customer churn and demonstrated how machine-learning models can support customer-retention decisions.
View Report