Uncovering of interesting structures in bank loan data through Bayesian networks

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60076658%3A12510%2F15%3A43890322" target="_blank" >RIV/60076658:12510/15:43890322 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Uncovering of interesting structures in bank loan data through Bayesian networks
Original language description
Given the explosive growth of data collected from current business environment, data mining methods can potentially discover new business knowledge to improve managerial decision. In This paper we use relatively novel data mining approach that employs discrete Bayesian network methodology to discover knowledge from data. More concretely we try to uncover structure and relationship among some socio-economic characteristics gathered by bank institution and credit risk in some sample of credit applicants.All data were collected during the 2013 - 2014. For this purpose we used two algorithms for Bayesian network structure discovering, namely hill climbing and growth-shrinking algorithms with a priori assigned relationship among some subset of socio-economics variables. The resulting structure was compared to others to BN. As a best BN model was identified structure without any implied restriction and derived by hill-climbing algorithm. BIC score for our best model was -13147.55. This "bes
Czech name
—
Czech description
—

Project
—
Continuities
I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year
2015
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
ICABR 2015, X. International Conference on Applied Business Research
ISBN
978-80-7509-379-0
ISSN
—
e-ISSN
—
Number of pages
887
Pages from-to
881
Publisher name
Mendelova univerzita v Brně
Place of publication
Brno
Event location
Madrid
Event date
Sep 14, 2015
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—

Similar results(10)