Benchmarking state-of-the-art classification algorithms for credit scoring

Baesens, Bart; Gestel, Tony Van; Viaene, Stijn; Stepanova, Maria; Suykens, Johan A. K.; Vanthienen, Jan

doi:10.1057/palgrave.jors.2601545

Cited by 717 publications

(509 citation statements)

References 19 publications

Supporting

Mentioning

447

Contrasting

Unclassified

Order By: Relevance

“…For the one-class k-NN classifier the number of neighbours was set to 10. This figure was selected in keeping with [11], who used 10-NN two-class classifier. The one-class Naïve Parzen classifier required no parameter tuning.…”

Section: Methodsmentioning

confidence: 99%

“…Many classification techniques have been used for credit scoring [11], some of which include traditional statistical methods such as logistic regression; non-parametric statistical methods, such as k-nearest neighbour; and sophisticated methods such as neural networks.…”

Section: Credit Scoringmentioning

confidence: 99%

“…The following two-class classifiers were assessed: (i) Logistic regression 4 ; (ii) Naïve Bayes [13] 4 ; (iii) Artificial neural network using a multilayer perceptron (MLP) 5 ; and (iv) Support Vector Machines [9] as they have been shown to perform well when applied to credit scoring problems in the past [11]. For all of the two-class classifiers a cut-off score is applied to the classifier output score, data instances above the cut-off are assigned to the positive class and those with scores below to the negative class.…”

Section: Classification Techniquesmentioning

confidence: 99%

“…We use Sensitivity, Specificity, as used by [11], and the harmonic mean of both of these scores to measure the classification quality of all classifiers used in our study. Sensitivity is calculated as: T P T P+FN and measures the proportion of positive (accept) examples that are predicted to be positive.…”

Section: Assessment Measuresmentioning

confidence: 99%

“…Two customers with similar characteristics can easily belong to different classes [31]. Also, credit scoring datasets are typically very noisy [11], particularly the Australian dataset [32]. Figures 2, 3 and 4 display the test set harmonic mean of all 10 classifiers for each of the datasets.…”

Section: Australianmentioning

confidence: 99%

See 4 more Smart Citations

Learning without Default: A Study of One-Class Classification and the Low-Default Portfolio Problem

Kennedy¹,

Namee²,

Delany³

2010

Artificial Intelligence and Cognitive Science

View full text Add to dashboard Cite

Abstract. This paper asks at what level of class imbalance one-class classifiers outperform two-class classifiers in credit scoring problems in which class imbalance, referred to as the low-default portfolio problem, is a serious issue. The question is answered by comparing the performance of a variety of one-class and two-class classifiers on a selection of credit scoring datasets as the class imbalance is manipulated. We also include random oversampling as this is one of the most common approaches to addressing class imbalance. This study analyses the suitability and performance of recognised two-class classifiers and one-class classifiers. Based on our study we conclude that the performance of the twoclass classifiers deteriorates proportionally to the level of class imbalance. The two-class classifiers outperform one-class classifiers with class imbalance levels down as far as 15% (i.e. the imbalance ratio of minority class to majority class is 15:85). The one-class classifiers, whose performance remains unvaried throughout, are preferred when the minority class constitutes approximately 2% or less of the data. Between an imbalance of 2% to 15% the results are not as conclusive. These results show that one-class classifiers could potentially be used as a solution to the low-default portfolio problem experienced in the credit scoring domain.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Credit Scoringmentioning

confidence: 99%

Section: Classification Techniquesmentioning

confidence: 99%

Section: Assessment Measuresmentioning

confidence: 99%

Section: Australianmentioning

confidence: 99%

See 3 more Smart Citations

Learning without Default: A Study of One-Class Classification and the Low-Default Portfolio Problem

Kennedy¹,

Namee²,

Delany³

2010

Artificial Intelligence and Cognitive Science

View full text Add to dashboard Cite

show abstract

Bayesian Network Classifiers

Goldszmidt¹

2011

Wiley Encyclopedia of Operations Research and Management Science

View full text Add to dashboard Cite

Recent work in supervised learning has shown that a surprisingly simple Bayesian classifier with strong assumptions of independence among features, called naive Bayes, is competitive with state-of-the-art classifiers such as C4.5. This fact raises the question of whether a classifier with less restrictive assumptions can perform even better. In this paper we evaluate approaches for inducing classifiers from data, based on the theory of learning Bayesian networks. These networks are factored representations of probability distributions that generalize the naive Bayesian classifier and explicitly represent statements about independence. Among these approaches we single out a method we call Tree Augmented Naive Bayes (TAN), which outperforms naive Bayes, yet at the same time maintains the computational simplicity (no search involved) and robustness that characterize naive Bayes. We experimentally tested these approaches, using problems from the University of California at Irvine repository, and compared them to C4.5, naive Bayes, and wrapper methods for feature selection.

show abstract