Τμηματοποίηση πελατών και πρόβλεψη αποχώρησης : μελέτη περίπτωσης ελληνικής εταιρείας διαδικτυακής παραγγελίας και διανομής προϊόντων

Χατζηνικολάου, Μιχαήλ

Customer segmentation and churn prediction for a greek online delivery app

Master Thesis

Author

Χατζηνικολάου, Μιχαήλ

Date

2025-10

Abstract

In today’s era, where businesses are required to manage large volumes of complex data, their effective utilization constitutes a critical factor in maintaining a competitive advantage. This thesis examines the case of a company operating in the online product delivery sector (food, supermarket goods, etc.), utilizing real customer data. Initially, data cleaning and exploratory data analysis are carried out with the aim of understanding the main behavioral patterns as well as the characteristics of the customer base (such as the most popular and most profitable product categories, the use of discount offers, the geographical distribution of consumers, etc.). Subsequently, two customer-centric approaches are applied: (a) Customer Segmentation using the RFM (Recency, Frequency, Monetary) methodology, to identify distinct groups with common characteristics for targeted marketing actions and broader business decisions, and (b) Churn Prediction using machine learning models (Logistic Regression, Random Forest, XGBoost). To address the class imbalance observed in the dataset, resampling techniques (SMOTE Oversampling, Random Undersampling) are applied. Finally, the models are evaluated and compared using appropriate metrics, with the aim of selecting the one most suitable for business application.

Postgraduate Studies Programme

Κυβερνοασφάλεια και Επιστήμη Δεδομένων

Department

Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Πληροφορικής

Number of pages

Language

Greek

URI

https://dione.lib.unipi.gr/xmlui/handle/unipi/18417

Collections

Τμήμα Πληροφορικής

Show full item record

Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση 3.0 Ελλάδα