Pregled bibliografske jedinice broj: 259267
A neural network classification of credit applicants in consumer credit scoring
A neural network classification of credit applicants in consumer credit scoring // Proceedings of the 24th IASTED International Multi-Conference « ; Artificial intelligence and applications» ; / Devedzic, Vladan (ur.).
Innsbruck: ACTA Press, 2006. str. 205-210 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 259267 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
A neural network classification of credit applicants in consumer credit scoring
Autori
Šarlija, Nataša ; Benšić, Mirta, Zekić-Sušac, Marijana
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings of the 24th IASTED International Multi-Conference « ; Artificial intelligence and applications» ;
/ Devedzic, Vladan - Innsbruck : ACTA Press, 2006, 205-210
Skup
24th IASTED International Multi-Conference « ; Artificial intelligence and applications» ;
Mjesto i datum
Innsbruck, Austrija, 13.02.2006. - 16.02.2006
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
credit scoring modeling; logistic regression; neural networks; radial basis function network
Sažetak
The paper aims to find an efficient model for consumer credit scoring using neural networks in comparison with logistic regression. A specific characteristic of the examined dataset was that the credit repayment period was not completed, assuming the existence of "good", "bad", and indeterminate ("poor") applicants which influenced the model accuracy. Five different modeling strategies were tested: (1) multinomial model with three categories of applicants, (2) binomial model using only good and bad applicants, (3) binomial model including poor applicants as good, (4) binomial model including poor applicants as bad, and (5) binomial model in which poor credit applicants were estimated by model 2 and then included in the dataset. The radial basis function network with softmax activation function produced best results among the three neural network algorithms tested. The results suggest that the best strategy to deal with poor applicants is to estimate them as good and bad, and then include into the model or to exclude them from the data set, although some further investigation is to be followed.
Izvorni jezik
Engleski
Znanstvena područja
Matematika, Ekonomija, Informacijske i komunikacijske znanosti
Napomena
Zbornik s konferencije je citiran u bazama: INSPEC, ISI Thomson, Elsevier (Engineering Information), Cambridge Scientific Abstracts, and Emerald
POVEZANOST RADA
Projekti:
0235002
Ustanove:
Ekonomski fakultet, Osijek,
Sveučilište u Osijeku, Odjel za matematiku
