Search results

1

A critical comparison of discriminant analysis and svm-based approaches to credit scoring

100%

Stąpor K.

Studia Ekonomiczne

|

2016

|

vol. 288

59-70

EN

Credit scoring models are the basis for financial institutions like retail and consumer credit banks. The purpose of these models is to evaluate the likelihood of credit applicants defaulting in order to decide whether to grant them credit. The paper compares two methodologies for building credit scoring models: heteroscedastic discriminant analysis-based with the support vector machines. The real-world credit dataset is used for comparison.

PL

Modele oceny ryzyka kredytowego stanowią podstawę działalności większości instytucji finansowych, zajmujących się udzielaniem kredytów. Celem takich modeli jest ewaluacja prawdopodobieństwa zaprzestania przez kredytobiorcę spłaty udzielonego mu kredytu. W artykule dokonano porównania dwóch modeli oceny ryzyka kredytowego, które wykorzystują nowe metody statystyczne, a także metody uczenia maszynowego do ich konstrukcji: heteroscedastyczną analizę dyskryminacyjną oraz maszyny wektorów podpierających. Dla dokonania porównania tych metod wykorzystany został ogólnie dostępny, niemiecki zbiór kredytowy.

2

APPLICATION OF SPARSE LINEAR DISCRIMINANT ANALYSIS FOR PREDICTION OF PROTEIN-PROTEIN INTERACTIONS

63%

Stąpor K., Fabian P.

Information Systems in Management

|

2016

|

vol. 5

|

issue 1

109-118

EN

To understand the complex cellular mechanisms involved in a biological system, it is necessary to study protein-protein interactions (PPIs) at the molecular level, in which prediction of PPIs plays a significant role. In this paper we propose a new classification approach based on the sparse discriminant analysis [10] to predict obligate (permanent) and non-obligate (transient) protein-protein interactions. The sparse discriminant analysis [10] circumvents the limitations of the classical discriminant analysis [4, 9] in the high dimensional low sample size settings by in-corporating inherently the feature selection into the optimization procedure. To characterize properties of protein interaction, we proposed to use the binding free energies. The performance of our proposed classifier is 75% ± 5%.

3

Heteroscedastic Discriminant Analysis Combined with Feature Selection for Credit Scoring

51%

Stąpor K., Smolarczyk T., Fabian P.

Statistics in Transition new series

|

2016

|

vol. 17

|

issue 2

265-280

EN

Credit granting is a fundamental question and one of the most complex tasks that every credit institution is faced with. Typically, credit scoring databases are often large and characterized by redundant and irrelevant features. An effective classification model will objectively help managers instead of intuitive experience. This study proposes an approach for building a credit scoring model based on the combination of heteroscedastic extension (Loog, Duin, 2002) of classical Fisher Linear Discriminant Analysis (Fisher, 1936, Krzyśko, 1990) and a feature selection algorithm that retains sufficient information for classification purpose. We have tested five feature subset selection algorithms: two filters and three wrappers. To evaluate the accuracy of the proposed credit scoring model and to compare it with the existing approaches we have used the German credit data set from the study (Chen, Li, 2010). The results of our study suggest that the proposed hybrid approach is an effective and promising method for building credit scoring models.

Refine search results

1 Information Systems in Management

1 Statistics in Transition new series

1 Studia Ekonomiczne

3 2016

3 Stąpor K.

2 Fabian P.

1 Smolarczyk T.

A critical comparison of discriminant analysis and svm-based approaches to credit scoring

APPLICATION OF SPARSE LINEAR DISCRIMINANT ANALYSIS FOR PREDICTION OF PROTEIN-PROTEIN INTERACTIONS

Heteroscedastic Discriminant Analysis Combined with Feature Selection for Credit Scoring