Επιχειρηματική αναλυτική στην πράξη : συσταδοποίηση και πρόβλεψη διαφυγόντων καταστημάτων για μια ηλεκτρονική πλατφόρμα διανομής φαγητού

Γιαννάκης, Γεώργιος

Business analytics in action : clustering and store churn prediction for an online food delivery platform

Master Thesis

Author

Γιαννάκης, Γεώργιος

Date

2023-10

Abstract

This thesis delves into the areas of Business Analytics (BA) and Business Intelligence (BI) with an emphasis on the application of Customer Analytics (CA) and Machine Learning (ML) techniques. The main aim is to analyze the advantages, challenges and techniques associated with BA and CA, while exploring the practical applications of store/customer clustering and churn prediction. Chapter 1 serves as an introduction, providing an overview of the research to follow. It also describes the purpose of the thesis, which is to explore the field of BA/BI and its applications, and more specifically in the context of CA. The chapter concludes with a brief description of the structure of the thesis. Chapter 2 delves into the fundamentals of BA and BI. We begin by introducing Business Analytics, discussing its methodologies, and exploring various perspectives within the field. In the rest of the chapter, we define Business Intelligence, emphasizing its importance in arriving at and making informed decisions. In addition, the historical development and distinctions between BA and BI are examined. In conclusion, we explore business value creation and identify current growth trends in the BA/BI sectors. Subsequently, Chapter 3 focuses specifically on BA. Here different BA methods will be explored, highlighting their advantages and challenges. This chapter will also develop the various fields of application of BA and present various variations of applications within different domains. Finally, we will cover different types of BA, including descriptive analytics, diagnostic analytics, predictive analytics, and prescriptive analytics. Chapter 4 develops Customer Analytics, framing its development in the e-commerce industry. Here we will study different market models, their types and differences. Next, we will look at the historical development of CA and deal with its advantages and techniques. Particular attention is paid to the clustering technique, with a detailed examination of various forms such as distribution-based clustering, density-based clustering, hierarchical clustering and centroid clustering. Additionally, the chapter explores pattern discovery and churn prediction. Chapter 5 serves as the practical part of the thesis. It begins with an introduction to the implementation tools used for the task. The chapter then focuses on data preparation, including data reading, data cleaning/organization, and Exploratory Data Analysis (EDA). Then, the applied clustering model is presented and we continue with the attempt to predict churn predictions, using techniques such as Decision Trees and Logistic Regression. Finally, the chapter concludes with business propositions arising from our analytical findings. Overall, this thesis provides a comprehensive exploration of BA, BI, and CA, with a special focus on store/customer clustering and churn prediction. Combining theoretical analysis and practical application, it contributes to the understanding of these areas and offers valuable insights for businesses seeking to leverage data-driven decision making.

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/16261
http://dx.doi.org/10.26267/unipi_dione/3683

Collections

Τμήμα Πληροφορικής

Show full item record

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