Selection of Support Vector Machines based classifiers for credit risk domain

A hybrid data mining model of feature selection algorithms and ensemble learning classifiers for credit scoring

“…Vukovic et al (2012) used the preference theory functions in the casebased reasoning (CBR) model for credit scoring model. Danenas and Garsva (2015) applied particle swarm optimization (PSO) for the optimal linear SVM classifier selection in the domain of credit risk.…”

Demonstrating non-inferiority of easy interpretable methods for insolvency prediction

“…A German and a Barbadian credit scoring data set with 300 and 503 non-creditworthy and 700 and 21,117 creditworthy applicants were used. Danenas et al (Danenas & Garsva, 2015) proposed an approach using linear SVMs also to safe computational complexity in the credit risk domain. Their evaluation was based on 5,527 risky and 15,961 non-risky entries.…”

Bank note authentication using decision tree rules and machine learning techniques

“…Danenas et al applied linear SVM for credit risk evaluation to show that it performs better than logistic regression and RBF network in terms of accuracy and identification of each class [16]. Boyacioglu et al applied MLP, self-organizing map (SOM), learning vector quantization(LVQ), competitive learning, multivariate discriminant, k-means cluster and logistic regression for predicting bank financial failures and concluded that MLP and LVQ performs well in terms of accuracy [17].…”