Analysing Unhealthy Shopping Patterns at ASDA Morley: Consumer Profiling Based on Basket Types and Geographical Distribution

Code: Jupyter notebook
Presentation: Full Report

Description: For my dissertation, I collaborated with ASDA to analyse health-conscious shopping behaviours using a dataset of over one million transactions from their Morley store. Utilising K-means clustering, I identified a customer segment characterized by low expenditure and minimal item purchases, who predominantly buy unhealthy products. This study aimed to provide ASDA with deeper insights into customer shopping patterns and to explore the relationships between demographic factors, such as the Index of Multiple Deprivation (IMD), Physical Food Environment (PPFI), and education levels. This project showcased my ability to manage large datasets and apply machine learning techniques for strategic analysis.
Results: The analysis revealed that customers with lower spending and fewer items in their baskets tend to choose unhealthier products. Specifically, residents of the postcode district LS10 exhibited less healthy shopping habits compared to those in six other postcode districts in Leeds. These findings helped ASDA guide customers in making healthier purchasing decisions.

Skills: Data Cleaning and Processing, Statistical Analysis, Hypothesis Testing, Machine Learning, Geospatial Analysis, Visualisation, Large-scale Data Handling.
Technology: Python, Pandas, Numpy, Statsmodels, Scipy, scikit-learn, K-means, Geopandas, Pyproj, Seaborn, Matplotlib.