Banca de DEFESA: Monique Lohane Xavier Silva

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : Monique Lohane Xavier Silva
DATE: 03/11/2022
TIME: 09:00
LOCAL: https://teams.microsoft.com/l/meetup-join/19%3aHuZvkaQInSrtD1bN2mVk5H1titThS94Z-uE8Rc4Oqi81%40threa
TITLE:

A Bayesian credit risk model for classification of default customers

KEY WORDS:

Risk score. Logistic regression. Bayesian inference.  Prior information.

PAGES: 56
BIG AREA: Ciências Exatas e da Terra
AREA: Probabilidade e Estatística
SUMMARY:

The aim of this work was to propose a bayesian credit risk model for classifying customers in terms of their default risk. The differential of the proposed methodology is the possibility of incorporating a priori information in the customer classification process and not just in the estimation of the customers’ evaluation parameters. The main advantage of this procedure is due to the simplicity in incorporating the expert’s opinion in the classification process, something that does not occur in traditional bayesian modeling, whose a priori information falls on the parameters of the models, which are usually abstract quantities and/or associated with covariates with multicollinearity problems. To illustrate the proposed methodology, a dataset in the literature was used and the results obtained showed that the model is useful for classifying customers in terms of their probability of default.

BANKING MEMBERS:
Presidente - 1373189 - EDUARDO YOSHIO NAKANO
Interno - 3000126 - JOSE AUGUSTO FIORUCCI
Externo à Instituição - MARCELO ANGELO CIRILLO - UFLA