Alessandra Borzomi'
Machine Learning Approaches for Customer Lifetime Value Prediction: A Case Study on Telepass Clients.
Rel. Paolo Garza. Politecnico di Torino, Corso di laurea magistrale in Data Science And Engineering, 2026
Abstract
Customer Lifetime Value (CLV) is a key metric for strategic marketing and resource allocation, as it provides a forward-looking estimate of the potential profitability of individual customers. This thesis investigates machine learning-based approaches for predicting CLV in a contractual business setting, using behavioral and transactional data from Telepass. In contrast to traditional CLV models based on static assumptions and aggregate-level estimates, the proposed approach adopts a modular, data-driven framework that estimates CLV at the individual customer level by modeling its two fundamental components: customer retention and profitability. The methodology combines customer churn prediction with margin estimation, leveraging both classical statistical techniques and modern machine learning models.
In particular, survival analysis methods, including the Cox Proportional Hazards model, are compared with tree-based approaches such as XGBoost
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Informazioni aggiuntive
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
