Procjena rizika preuranjenog vraćanja kredita putem strojnog učenja

undergraduate thesis

• Title (english): Machine Learning Based Early-return Credit Risk Assesment
• Author: Ivona Mramor
• Mentor: Vladimir Čeperić (mentor)
• Committee member: Vladimir Čeperić (predsjednik povjerenstva)
• Committee member: Mladen Karan (član povjerenstva)
• Committee member: Igor Stančin (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Banke se često nalaze u nedoumici prilikom sklapanja ugovora o kreditu jer ne mogu znati hoće li se klijent držati ugovora, zbog toga banke razvijaju razne strategije kako bi mogle odrediti rizike i jedan od rizika je mogućnost da klijent vrati kredit prije dogovorenog ro­ka.U ovom radu opisan pristup određivanju rizika preuranjenog vraćanja kredita. Opisani su postupci obrade skupa podataka. Upoznavanje s algoritmom XGBoost-a i svim nje­govim značajkama koji ga dovode u prednost naspram ostalih algoritama strojnog učenja. Izgradnja inicijalnog modela za procjenu rizika preuranjenog vraćanja kredita te detaljni opis koraka koje je potrebno izvesti kako bi se parametri XGBoost-a podesili na vrijednosti koji će poboljšati točnost modela.

Abstract (english)

Banks often find themselves in doubt when concluding a loan agreement because they can not know if a client will behave as it was agreed in contracts, so banks develop various strategies to determine risks and one of the risks is if the client pay off their loan earlier. In this paper is described approach to determining the risk of early repayment of the loan. Also, the data set processing procedures are described. Acquainted with the XGBoost al­gorithm and all its features that take precedence over other machine learning algorithms.It is built an initial model for estimating the risk of premature loan repayment and a detailed description of the steps that need to be performed to set the XGBoost parameters to a value that will improve the accuracy of the model.

• Keywords: klasifikacija; pojačavanje; treniranje; prenaučenost; testiranje; validacija; XGBoost; stabla odluke
• Keywords (english): Classification; Boosting; Overfitting; Training; Testing; Validation:XGBoost; Decision Tree
• Language: croatian
• URN:NBN: urn:nbn:hr:168:869145
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2020-01-07 20:38:02