Data Depth Based Support Vector Machines for Predicting Corporate Bankruptcy

摘要

Disclosed is a method for predicting company bankruptcy by using a data depth. According to an embodiment of the present invention, the method for predicting company bankruptcy by using a data depth reduces financial statements in one dimension by using a data depth as a measure since the many financial statements do not follow multivariate normal distribution. Also, the method for predicting company bankruptcy by using a data depth displays the financial statements in two dimension by a graphic method which is DD-plot and classifies the displayed financial statement by a support vector machine. The method for predicting company bankruptcy can improve a predicted company bankruptcy probability by using a result in which the displayed financial statement is classified.