Bankruptcy Classification of Firms Investigated by the US Securities and Exchange Commission: An Evolutionary Ensemble Computing Model Approach
International Journal of Applied Decision Sciences, 2009, Vol.2, No.4, pp.360-388.
42 Pages Posted: 31 Aug 2009 Last revised: 20 Feb 2018
Date Written: August 26, 2009
This paper develops an adaptive ensemble model for bankruptcy classification of firms cited in the SEC's Accounting and Auditing Enforcement Releases (AAER). We develop a Genetic Algorithm (GA) model for bankruptcy classification of AAER firms. Our research contributes to the bankruptcy literature in several ways. First of all, it fills a gap in the bankruptcy literature by developing a domain specific model for AAER firms. Secondly, by using financial and non-financial variables, the GA model generates and optimizes a set of 'if-then' comprehensible rules for the financial failure classification of AAER firms. A Genetic Algorithm model can provide a greater degree of accuracy in predicting financial failure of firms than classical statistical models. Thirdly, we develop a model using bagging that incorporates the output from different models or sources. Finally, we demonstrate the key role of the fitness function in determining the successful performance of a financial failure GA model.
Keywords: SEC, AAER, genetic algorithm, evolutionary computing, fitness function, concept learning, bagging, bankruptcy classification, ensemble, multiple classifier, hybrid classifier, accounting and political economy
JEL Classification: C45, C61, G18, G38, K22, L81, M40
Suggested Citation: Suggested Citation