Logo Lanfrica

ndeyekwade-cmd/Data-Preprocessing-AIMS

Domain:

socioeconomic

Record type:

projectdataset
Creator:
nde
Host:
Mobile Money Transaction Intelligence across East Africa - AIMS Senegal Academic Project # Data Preprocessing - Mobile Money Transaction Intelligence **AIMS Senegal Academic Project** *Master's in Mathematical Sciences with Big Data Specialization* --- ## 📋 Project Overview This project analyzes **Mobile Money transaction patterns across East Africa** (Kenya, Tanzania, and Uganda) to understand digital financial behavior, identify trends, and provide strategic recommendations for improving financial inclusion and service optimization. ### Team Members - **Ndeye Khady Wade** - **Maimouna Ndoye** - **Soukeyna Touré** --- ## 🎯 Project Objectives ### Strategic Goals 1. **Time Pattern Analysis** - Identify peaks and troughs in transaction activity 2. **User Behavior Analysis** - Understand transaction habits and demographics 3. **Geographic Disparities** - Compare rural vs. urban financial inclusion 4. **Competitive Analysis** - Evaluate performance across providers and countries --- ## 📊 Dataset Overview ### Dataset Characteristics - **Size:** 5,000 transactions - **Coverage:** Kenya, Tanzania, Uganda - **Time Period:** July - December 2023 - **Missing Values:** 0.27% - **Final Sample:** 5,000 observations ### Key Variables | Variable | Description | |----------|-------------| | `transaction_id` | Unique transaction identifier | | `timestamp` | Date and time of transaction | | `country` | Country (Kenya/Tanzania/Uganda) | | `provider` | Mobile Money provider | | `transaction_type` | Type (deposit/withdrawal/payment) | | `user_type` | User category (individual/business/agent) | | `amount_usd` | Transaction amount in USD | | `user_age` | User age group | | `is_rural` | Rural/Urban indicator | --- ## 🔍 Analysis Framework ### 1. Exploratory Data Analysis - **Transaction Volume by Hour** - Peak activity: 8 AM - 6 PM (~380 transactions/hour) - Low activity: Midnight - 6 AM - **Monthly Distribution** - Peak: August (890 transactions) - Trough: September, November - Recovery: December - **Weekday vs Weekend Patterns** - Stable amounts: USD 20-30 - …