دانلود رایگان مقاله انگلیسی بررسی سیستماتیک مدل پیش بینی ورشکستگی: انتخاب ابزاری - الزویر 2017

abstract

The bankruptcy prediction research domain continues to evolve with many new different predictive models developed using various tools. Yet many of the tools are used with the wrong data conditions or for the wrong situation. Using the Web of Science, Business Source Complete and Engineering Village databases, a systematic review of 49 journal articles published between 2010 and 2015 was carried out. This review shows how eight popular and promising tools perform based on 13 key criteria within the bankruptcy prediction models research area. These tools include two statistical tools: multiple discriminant analysis and Logistic regression; and six artificial intelligence tools: artificial neural network, support vector machines, rough sets, case based reasoning, decision tree and genetic algorithm. The 13 criteria identified include accuracy, result transparency, fully deterministic output, data size capability, data dispersion, variable selection method required, variable types applicable, and more. Overall, it was found that no single tool is predominantly better than other tools in relation to the 13 identified criteria. A tabular and a diagrammatic framework are provided as guidelines for the selection of tools that best fit different situations. It is concluded that an overall better performance model can only be found by informed integration of tools to form a hybrid model. This paper contributes towards a thorough understanding of the features of the tools used to develop bankruptcy prediction models and their related shortcomings.

Conclusion

The bankruptcy prediction research domain continues to evolve with many new models developed using various tools. Yet many of the tools are used with the wrong data conditions or for the wrong situation. This study used a systematic review, to reveal how eight popular and promising tools (MDA, LR, ANN, SVM, RS, CBR, DT and GA) perform with regard to various important criteria in the bankruptcy prediction models (BPM) study area. Overall, it can be concluded that there is no singular tool that is predominantly better than all other tools in relation to all identified criteria. It is however clear that each tool has its strengths and weaknesses that make it more suited to certain situations (i.e. data characteristics, developer aim, among others) than others. The framework presented in this study clearly provides a platform that allows a well-informed selection of tool(s) that can best fit the situation of a model developer. The implication of this study is that BPM developers can now make an informed decision when selecting a tool for their model rather than make selection based on popularity or other unscholarly factors. In essence, tools will be more regularly selected based on their strength. Another implication is that BPMs with better performance with regards to end users’ requirement will be more commonly developed. This is better than to continue with the present trend of ‘one size fits all’ where a BPM tool is assumed to be good enough for the very different users/clients (e.g. financiers, clients, owners, government agencies, auditors, among others) that need them. The framework in this study will also reduce the timewasting process of developing many BPMs with different tools in order to select the best after a series of test; only the tools that best fit a developer’s situation will be used and compared.