Majority Voting Combination of Multiple Classifiers for Bankruptcy Prediction

The problem of bankruptcy prediction is one of the most actively studied nowadays. Many studies have been conducted on corporate bankruptcy prediction using data mining techniques. Artificial Neural Networks, Support Vector Machine and Decision Tree Algorithms are three current methods for data mining to prediction bankruptcy. This study puts forward a bankruptcy prediction methods based on majority voting combination of Artificial Neural Networks, Support Vector Machine and Decision Tree. Statistical population of this study includes 126 sound companies and 126 bankrupt companies, active in Tehran Stock Exchange Market between 2005 and 2011, which were studied for the three years of t , t-1 and t-2 . The results show that combination of relative majority voting with 92.20% accuracy in the year t and 88.88% accuracy in the year t-1 and 80.22% accuracy in the year t-2 is able to prediction corporate bankruptcy

Keywords:

Bankruptcy , Artificial Neural Networks , Support Vector Machine , Decision Tree , Majority Voting Combination