Institutional Repository of Vilnius University
A comparison of machine learning upsell propensity models for the telecommunication sector /
Abstract
dc:descriptionIn the telecommunication industry, predicting which service is most valued by the customer could help personalize service bundles and offers, thus increasing both customer satisfaction and revenue. One way of predicting the importance of a service to a customer is by modelling upsell propensity with machine learning algorithms. This study aims to find the best method to predict upsell and to analyse the sufficiency of a dataset from a telecommunication company containing customer account and billing data. Four models - a Decision Tree, Random Forest, Logistic Regression and Neural Network are compared and their results and performances are analysed. The results show that the best method for predicting upsell with in this project available data, both computation and performance-wise, is the decision tree, although all models achieved a similar accuracy of about 79$\%$. By analysing the results, it was also found that valuable insights on service importance can be made in only about a third of cases where customers have multiple services.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2021
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Skerath, Lina,
- Contributors dc:contributor
-
- Šimėnas, Raivydas
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD107118758&prefLang=en_US