Marijana Zekić-Sušac

Modelling small-business credit scoring by using logistic regression, neural networks and decision trees

Previous research on credit scoring that used statistical and intelligent methods was mostly focused on commercial and consumer lending. The main purpose of this paper is to extract important features for credit scoring in small-business lending on a dataset with specific transitional economic conditions using a relatively small dataset. To do this, we compare the accuracy of the best models extracted by different methodologies, such as logistic regression, neural networks (NNs), and CART decision trees. Four different NN algorithms are tested, including backpropagation, radial basis function network, probabilistic and learning vector quantization, by using the forward nonlinear variable selection strategy. Although the test of differences in proportion and McNemars test do not show a statistically significant difference in the models tested, the probabilistic NN model produces the highest hit rate and the lowest type I error. According to the measures of association, the best NN model also shows the highest degree of association with the data, and it yields the lowest total relative cost of misclassification for all scenarios examined. The best model extracts a set of important features for small-business credit scoring for the observed sample, emphasizing credit programme characteristics, as well as entrepreneurs personal and business characteristics as the most important ones. Copyright

Small business credit scoring: a comparison of logistic regression, neural network, and decision tree models

The paper compares the models for small business credit scoring developed by logistic regression, neural networks, and CART decision trees on a Croatian bank dataset. The models obtained by all three methodologies were estimated; then validated on the same hold-out sample, and their performance is compared. There is an evident significant difference among the best neural network model, decision tree model, and logistic regression model. The most successful neural network model was obtained by the probabilistic algorithm. The best model extracted the most important features for small business credit scoring from the observed data

Prediction of influenza vaccination outcome by neural networks and logistic regression

The major challenge in influenza vaccination is to predict vaccine efficacy. The purpose of this study was to design a model to enable successful prediction of the outcome of influenza vaccination based on real historical medical data. A non-linear neural network approach was used, and its performance compared to logistic regression. The three neural network algorithms were tested: multilayer perceptron, radial basis and probabilistic in conjunction with parameter optimization and regularization techniques in order to create an influenza vaccination model that could be used for prediction purposes in the medical practice of primary health care physicians, where the vaccine is usually dispensed. The selection of input variables was based on a model of the vaccine strain which has frequently been changed and on which a poor influenza vaccine response is expected. The performance of models was measured by the average hit rate of negative and positive vaccine outcome. In order to test the generalization ability of the models, a 10-fold cross-validation procedure revealed that the model obtained by multilayer perceptron produced the highest average hit rate among neural network algorithms, and also outperformed the logistic regression model with regard to sensitivity and specificity. Sensitivity analysis was performed on the best model and the importance of input variables was discussed. Further research should focus on improving the performance of the model by combining neural networks with other intelligent methods in this field.

Leveraging the Strengths of Choice Models and Neural Networks: A Multiproduct Comparative Analysis*

Choice models and neural networks are two approaches used in modeling selection decisions. Defining model performance as the out-of-sample prediction power of a model, we test two hypotheses: (i) choice models and neural network models are equal in performance, and (ii) hybrid models consisting of a combination of choice and neural network models perform better than each stand-alone model. We perform statistical tests for two classes of linear and nonlinear hybrid models and compute the empirical integrated rank (EIR) indices to compare the overall performances of the models. We test the above hypotheses by using data for various brand and store choices for three consumer products. Extensive jackknifing and out-of-sample tests for four different model specifications are applied for increasing the external validity of the results. Our results show that using neural networks has a higher probability of resulting in a better performance. Our findings also indicate that hybrid models outperform stand-alone models, in that using hybrid models guarantee overall results equal or better than the two stand-alone models. The improvement is particularly significant in cases where neither of the two stand-alone models is very accurate in prediction, indicating that the proposed hybrid models may capture aspects of predictive accuracy that neither stand-alone model is capable of on their own. Our results are particularly important in brand management and customer relationship management, indicating that multiple technologies and mixture of technologies may yield more accurate and reliable outcomes than individual ones.

Comparison of intelligent systems in detecting a child's mathematical gift

This paper compares the efficiency of two intelligent methods: expert systems and neural networks, in detecting childrens mathematical gift at the fourth grade of elementary school. The input space for the expert system and the neural network model consisted of 60 variables describing five basic components of a childs mathematical gift identified in previous research. The expert system estimated a childs gift based on heuristically defined logic rules, while the scientifically confirmed psychological evaluation of gift based on Ravens standard progressive matrices was used at the output of neural network models. Three neural network algorithms were tested on a Croatian dataset. The results show that both the expert system and the neural network recognize more pupils as mathematically gifted than teachers do. The expert system produces the highest average hit rate, although the highest accuracy in classifying gifted children is obtained by the radial basis neural network algorithm, which also yields lower type II error. Due to the ability of expert systems to explain the result, it can be suggested that both the expert system and the neural network model have potential to serve as effective intelligent decision support tools in detecting mathematical gift in early stage, therefore enabling its further development.

Data Mining as Support to Knowledge Management in Marketing

Abstract Background: Previous research has shown success of data mining methods in marketing. However, their integration in a knowledge management system is still not investigated enough. Objectives: The purpose of this paper is to suggest an integration of two data mining techniques: neural networks and association rules in marketing modeling that could serve as an input to knowledge management and produce better marketing decisions. Methods/Approach: Association rules and artificial neural networks are combined in a data mining component to discover patterns and customers’ profiles in frequent item purchases. The results of data mining are used in a web-based knowledge management component to trigger ideas for new marketing strategies. The model is tested by an experimental research. Results: The results show that the suggested model could be efficiently used to recognize patterns in shopping behaviour and generate new marketing strategies. Conclusions: The scientific contribution lies in proposing an integrative data mining approach that could present support to knowledge management. The research could be useful to marketing and retail managers in improving the process of their decision making, as well as to researchers in the area of marketing modelling. Future studies should include more samples and other data mining techniques in order to test the model generalization ability.

A Comparison of Machine Learning Methods in a High-Dimensional Classification Problem

Abstract Background: Large-dimensional data modelling often relies on variable reduction methods in the pre-processing and in the post-processing stage. However, such a reduction usually provides less information and yields a lower accuracy of the model. Objectives: The aim of this paper is to assess the high-dimensional classification problem of recognizing entrepreneurial intentions of students by machine learning methods. Methods/Approach: Four methods were tested: artificial neural networks, CART classification trees, support vector machines, and k-nearest neighbour on the same dataset in order to compare their efficiency in the sense of classification accuracy. The performance of each method was compared on ten subsamples in a 10-fold cross-validation procedure in order to assess computing sensitivity and specificity of each model. Results: The artificial neural network model based on multilayer perceptron yielded a higher classification rate than the models produced by other methods. The pairwise t-test showed a statistical significance between the artificial neural network and the k-nearest neighbour model, while the difference among other methods was not statistically significant. Conclusions: Tested machine learning methods are able to learn fast and achieve high classification accuracy. However, further advancement can be assured by testing a few additional methodological refinements in machine learning methods.

Modeling children's mathematical gift by neural networks and logistic regression

The purpose of the paper was to extract important features of childrens mathematical gift by using neural networks and logistic regression, in order to create a model that will assist teachers in elementary schools to recognize mathematically gifted children in an early stage, therefore enabling further development and realization of that gift. The initial model was created on the basis of a theoretical background and heuristical knowledge on giftedness in mathematics, including five components: (1) mathematical competencies, (2) cognitive components of gift, (3) personal components that contribute gift development, (4) environmental factors, and (5) efficiency of active learning and exercising methods, as well as grades and out-of-school activities of pupils in the fourth year of elementary school. The three neural network classification algorithms were tested in order to extract the important variables for detecting mathematically gifted children. The best neural network model was selected on the basis of a 10-fold cross-validation procedure. The model was also investigated by the logistic regression. Important predictors detected by two methods were compared and analyzed. The results show that both methods extract similar set of variables as the most important, including grades in mathematics, mathematical competencies of a child regarding numbers and calculating, but also grades in the literature, and environmental factors.

Is Operational Research Attractive Enough at Higher Education Institutions in Croatia

In an age where information technology, the Internet, smart phones, mobile applications, interdisciplinarity, big data, business analytics, artificial intelligence, and the fourth industrial revolution have become prevalent, Operational Research (OR) as a scientific and academic discipline is also subject to a specific transformation and evolution. This paper covers the current situation with OR based, on the mentioned trends by analysing OR-related programmes and courses offered at university level, specifically, at higher education institutions (HEIs) in Croatia. Also, aspects of collaboration between HEIs and the business community relating to OR will be studied, as well as the role of the Croatian Operational Research Society (CRORS) as a community that gathers OR researchers and promotes OR as a scientific discipline (Croatian Operational Research Society 2017). Some suggestions to improve the current OR situation at Croatian HEIs, and its impact on economy are also given. The analysis was based on a survey sent to HEIs in Croatia, and subsequently organising a round table discussion and consultations with Croatian employers. The survey questions related to the courses and programmes offered at Croatian HEIs, the number of students attending the courses, recent innovations, collaboration with the business community and plans to make OR even more appealing. The results reveal that the number of OR-related courses is significant, though most are electives, and the lack of OR majors across all higher-education levels in Croatia. In general, OR courses are required to follow terminology trends and utilize more appealing terms such as data analytics and business analytics. Recent and more intensive collaboration with the business community which seeks experts in optimization, analytics, and other OR areas is becoming promising.

Elucidating clinical context of lymphopenia by nonlinear modelling

A nonlinear approach for detecting relative lymphopenia is suggested by using a health data record based on simple clinical parameters. Two classification methods, neural networks and decision trees, were applied to detect whether a patient has a positive or a negative lymphopenia outcome. Due to a large dimension of input space, a feature selection method was used in the pre-processing stage. All tested models were validated on the same out-of-sample dataset, and a 10-fold cross-validation procedure for testing generalization ability of the models was conducted. The models were compared according to their classification accuracy in the sense of the average hit rate, specificity and sensitivity. The results show that (1) the best neural network model slightly outperforms the decision tree model, (2) the reduced model provides even higher accuracy than the models with all available data, and (3) both methods similarly rank five important predictors of lymphopenia. The paper discusses the relevance of extracted features, and suggests some guidelines for further research.