Institutional Repository of Vilnius University
Kreditų rizikos prognozavimas analizuojant tarpusavio skolinimo platformos duomenis /
Abstract
dc:descriptionIn the bachelor’s thesis, the credit data of the peer-to-peer lending platform "SAVY" were analyzed by collecting it through web-scraping. After selecting the most important features and testing several quantitative variable scaling transformations, five machine learning algorithms were applied to classify credits into risky and non-risky categories. The accuracy metric was used for the binary classification task, and most models showed good predictive ability and could be used for forecasts. Changing the prediction threshold further improved the accuracy of identifying risky contracts and allowed for the correct identification of up to 92 % of risky loans.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Macijauskaitė, Sandra,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD192845155&prefLang=en_US