fanyicharllson/mobile-money-transaction-analysis

Machine learning pipeline for classifying mobile money users (MTN MoMo & Orange Money) into activity segments — CSC 3221 Final Project, ICT University Cameroon. # Mobile Money Transaction Analysis ## CSC 3221 — Introduction to Data Science | ICT University Cameroon --- ## Group Members | Name | Student ID | Role | |------|-----------|------| | FANYI CHARLLSON FANYI | ICTU20233841 | Data Collection Lead + EDA | | ETAN WILL JOHN | ICTU20233606 | Data Cleaning + Feature Engineering | | NJIFON ERIC DENIS | ICTU20234391 | Modeling + Model Evaluation | | MADONGUE JEANNE LESLINE | ICTU20222931 | Report Writing + Visualization | | MAKUETE TENETCHIT JULIE FORTUNE | ICTU20233770 | Report Writing | --- ## Project Title **Mobile Money Transaction Pattern Analysis — User Activity Segment Classification** ## Brief Description This project analyzes mobile money transaction patterns (MTN Mobile Money & Orange Money) among 60 users across 8 regions of Cameroon to classify users into Low, Medium, and High activity segments using machine learning. The best model (Random Forest) achieved 78.9% cross-validated accuracy, nearly 2× the 44.4% baseline. **Prediction target:** User Activity Segment (Low / Medium / High) — 3-class classification --- ## File Structure ``` DataScience_Final_Group57_[Team Zeta-Chi]/ │ ├── README.md ← This file ├── CONTRIBUTIONS.md ← Member contribution breakdown ├── requirements.txt ← Python dependencies │ ├── 1_Data_Collection/ │ ├── raw_data_anonymized.csv ← Full dataset (60 users × 2,930 transactions) │ ├── raw_demographics.csv ← Demographic data only │ ├── raw_transactions.csv ← Raw transaction records (with intentional issues) │ ├── consent_form.pdf ← Informed consent template used │ ├── questionnaire.pdf ← Data collection questionnaire │ └── data_collection_report.pdf ← Sampling strategy & quality assurance │ ├── 2_Data_Cleaning/ │ ├── data_cleaning.py ← Data cleaning notebook (Python script) │ ├── cleaned_data.csv ← Final cleaned dataset (mod …