A hybrid KMV model, random forests and rough set theory approach for credit rating

Yeh, Ching-Chiang; Lin, Fengyi; Hsu, Chih‐Yu

doi:10.1016/j.knosys.2012.04.004

Cited by 98 publications

(49 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The database provided by the credit bureau has 10,356 customer observations of direct retail consumer credit operations and 198 variables for the year 2014. Although random trees and BART were designed for larger datasets it is not unusual to find papers that aim to compare estimation methods designed for big datasets with sample sizes equivalent to ours, see, for instance, Chipman et al (2010), Yeh et al (2012), Leong (2016), Abellán & Castellano (2017), Bequé & Lessmann (2017), and several papers analysed in Lessmann et al (2015) review.…”

Section: Databasementioning

confidence: 99%

Application of bayesian additive regression trees in the development of credit scoring models in Brazil

Filho

Artes

2018

Prod.

View full text Add to dashboard Cite

Paper aims: This paper presents a comparison of the performances of the Bayesian additive regression trees (BART), Random Forest (RF) and the logistic regression model (LRM) for the development of credit scoring models. Originality:It is not usual the use of BART methodology for the analysis of credit scoring data. The database was provided by Serasa-Experian with information regarding direct retail consumer credit operations. The use of credit bureau variables is not usual in academic papers.Research method: Several models were adjusted and their performances were compared by using regular methods. Main findings:The analysis confirms the superiority of the BART model over the LRM for the analyzed data. RF was superior to LRM only for the balanced sample. The best-adjusted BART model was superior to RF. Implications for theory and practice:The paper suggests that the use of BART or RF may bring better results for credit scoring modelling. KeywordsCredit. Machine learning. Logistic regression. BART. Random Forest.

show abstract

Section: Databasementioning

confidence: 99%

Application of bayesian additive regression trees in the development of credit scoring models in Brazil

Filho

Artes

2018

Prod.

View full text Add to dashboard Cite

show abstract

“…Note, however, that the current literature on financial institutions still lacks a more systematic and integrated approach in the second stage that might put these predictive modelling techniques into perspective (Yeh et al ., ; Hajek and Michalak, ; Sun et al ., ). This paper aims to contribute to the current state of the art of the literature by combining artificial neural networks and TOPSIS in a two‐stage procedure to predict banking performance in Angola.…”

Section: Literature Reviewmentioning

confidence: 99%

Predicting Efficiency in Angolan Banks: A Two‐Stage TOPSIS and Neural Networks Approach

Barros

Macanda

2015

South African J Economics

View full text Add to dashboard Cite

This paper presents an efficiency assessment of the Angolan banks using Technique for Order Preference by Similarity to the Ideal Solution (TOPSIS). TOPSIS is a multi‐criteria decision‐making technique similar to data envelopment analysis, which ranks a finite set of units based on the minimisation of distance from an ideal point and the maximisation of distance from an anti‐ideal point. In this research, TOPSIS is used first in a two‐stage approach to assess the relative efficiency of Angolan banks using the most frequent indicators adopted by the literature. Then, in the second stage, neural networks are combined with TOPSIS results as part of an attempt to produce a model for banking performance with effective predictive ability. The results reveal that variables related to cost structure have a prominent negative impact on efficiency. Findings also indicate that the Angolan banking market would benefit from higher level of competition between institutions.

show abstract

“…(3) Hybrid comparison: This work also compares the performance of the similar hybrid models: the neighborhood RS (NRS) + SVM model [59]; the random forest (RDF) + RS model [60]; and the MEPA-RS model, which has the best accuracy of the two proposed models. These models are applied to classify credit ratings.…”

Section: Table 13mentioning

confidence: 99%

Hybrid models based on rough set classifiers for setting credit rating decision rules in the global banking industry

Chen

Cheng

2013

Knowledge-Based Systems

View full text Add to dashboard Cite

A hybrid KMV model, random forests and rough set theory approach for credit rating

Cited by 98 publications

References 56 publications

Application of bayesian additive regression trees in the development of credit scoring models in Brazil

Application of bayesian additive regression trees in the development of credit scoring models in Brazil

Predicting Efficiency in Angolan Banks: A Two‐Stage TOPSIS and Neural Networks Approach

Hybrid models based on rough set classifiers for setting credit rating decision rules in the global banking industry

Contact Info

Product

Resources

About