Bank_Analysis

🏦 Bank Service & Transaction Analysis

🎓 Project Context

This project was developed as a final course project for the Database and Data Mining course during the “Digital Pioneers Initiative - مبادرة الرواد الرقميون”, under the supervision of Dr. Mohamed Sobhy.

Project Presentation

https://bank-ai-show.lovable.app/

Live Demo

https://esraamahmoud09.github.io/Bank_Analysis/

📌 Project Overview

This project delivers a comprehensive, data-driven analysis of banking operations, focusing on customer demographics, transaction behavior, and service status optimization.

By combining Exploratory Data Analysis (EDA) and Machine Learning techniques, the analysis uncovers key drivers behind service outcomes and transaction patterns, helping improve decision-making and operational efficiency.

📊 Business Insights & Key Findings

1️⃣ Service Performance

Service Distribution: Bill Payment dominates service requests (~83%), followed by Loans and Insurance services.
Service Status Tracking: A large portion of services are still In Progress (42.5%), while Rejected services account for ~13.4%, indicating potential inefficiencies in processing or eligibility criteria.
Customer Demographics: The Senior age group shows the highest engagement with banking services compared to Young Adults and Middle-Aged customers.

2️⃣ Transaction Patterns

Transaction Medium: Cash is the most commonly used payment method (55.7%), followed by Credit Cards (25.4%).
Transaction Type Distribution: Balance Inquiries (42.3%) are the most frequent transactions, reflecting high customer interest in monitoring account activity.

3️⃣ Geographical Insights

The Chicago Branch is the main operational hub, handling approximately 75.5% of total activity, indicating a strong regional concentration.

🤖 Machine Learning Insights

🔍 Feature Importance (Random Forest Model)

The model identifies the most influential features affecting Service Status:

Service_Amount → Strongest predictor
Customer_Name & Age → Key demographic factors
Balance → Indicator of financial eligibility

📈 Correlation Analysis

A detailed Correlation Heatmap is used to:

Identify relationships between numerical and encoded variables
Detect and reduce multicollinearity
Highlight important features such as:
- Service_Amount_Log
- Transaction_Amount

🛠️ Tech Stack

Python → Core programming language
Pandas & NumPy → Data cleaning & feature engineering
Matplotlib & Seaborn → Data visualization & heatmaps
Scikit-learn → Machine learning modeling & evaluation

🎯 Project Objectives

Understand customer behavior and service usage patterns
Improve service approval and processing efficiency
Identify key factors influencing banking operations
Support data-driven decision-making in financial services

🚀 Future Enhancements

Deploy model as an API for real-time predictions
Build interactive dashboards using Power BI or Streamlit
Apply advanced models (XGBoost, Neural Networks)
Incorporate time-series analysis for transaction forecasting

👥 Team Members

Name	Responsibility	Role
Hayam Medhat Wahdan	Report Implementation + Bank System	Team Leader
Hoda Mohamed Ezzat	Data Integration + Database Implementation	Member
Nada Fahmy Fahmy	SQL Questions + ER Diagram	Member
Esraa Mohamed Attia	EDA + Python Scripts	Member
Esraa Shawky Mahmoud	EDA + Orange Data Mining	Member

📬 Contact

If you have any questions or suggestions, feel free to reach out or open an issue.

📧 Email: hayamm.wahdan@gmail.com
🔗 LinkedIn: Hayam Wahdan

📧 Email: huda.elhamahmy@gmail.com
🔗 LinkedIn: Hoda Elhamahmy

📧 Email: roaaezzawy@gmail.com
🔗 LinkedIn: Esraa Al-Azzawy

📧 Email: nada.fahmy2244@gmail.com
🔗 LinkedIn: Nada Fahmy

📧 Email: esraashawky09@gmail.com
🔗 LinkedIn: Esraa Mahmoud

✨ This project demonstrates practical application of data analysis and machine learning in the banking domain.

This site is open source. Improve this page.