Abstract
dc:descriptionClassification of the incoming payments to the private customers of the bank is essential for customer behaviour and income analysis. However, more often banks classify the outgoing transactions and only specific categories of customer incomes. This thesis proposes possible categories of incoming transactions and investigates how do different algorithms perform at classification. There were manually obtained 21 new class of customer incomes. The Support Vector Machine, Naïve Bayes and Gradient Boosting classification algorithms show promising results and can be implemented in the bank instead of rule-based approach. The main focus in the future work is to reduce a noise in the textual data by automatic correction of mistakes in the transaction description.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2021
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Kolesnikova, Julija,
- Contributors dc:contributor
-
- Kazakevičius, Vytautas
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:ELABAETD81590519&prefLang=en_US