This repository contains a professional collection of data mining projects and analytical labs. The work focuses on exploratory data analysis (EDA), statistical validation, and business intelligence discovery.
The flagship study in this portfolio (Lab 03) analyzes retail transaction patterns across six Moroccan cities to identify the "Sunday Effect."
- Premium basket value: Sunday transactions show a 14% higher average value (1,120 MAD) compared to weekdays. This suggests consumers shift toward planned, high-value purchases during the weekend.
- Tech-driven growth: High-tech products, such as smartphones and peripherals, generate approximately 32% of total Sunday revenue.
- Geographical consistency: The surge in smartphone sales is a robust national trend verified in all analyzed cities, including Casablanca, Rabat, and Tangier.
- lab-01-intro: Introduction to data mining concepts and methodology.
- lab-02-pandas-matplotlib: Advanced data manipulation and visualization using NumPy, Pandas, and Matplotlib.
- lab-03-sunday-effect: Retail analysis using statistical hypothesis testing and multi-city validation.
- Language: Python (Jupyter)
- Libraries: Pandas, NumPy, Scikit-learn
- Visualization: Matplotlib, Seaborn
Authored by Youssef Fellah. Developed for the Engineering Cycle at Mundiapolis University.