Prediction of Financial Distress, Using Metaheuristic Models

Abstract

Investors need to assess and analyze the financial statement, to make the logical decision. Using financial ratios is one of the most common methods. The main purpose of this research is to predict the financial crisis, using ratios of liquidity. Four models, Support vector machine, neural network back propagation, Decision trees and Adaptive Neuro-Fuzzy Inference System has been compared.Furthermore, the ratios of liquidity considered in a period of 89_93. The research method is qualitative and quantitative and type of casual comparative. The result indicates that the accuracy of the neural network, Decision tree, and Adaptive Neuro-Fuzzy Inference System showed that there is a significant differently 0/000 and 0/005 years this is more than support vector machine result. Therefore the result of support vector machine showed that there is a significant differently 0/001 in years. This has been shown that neural network in 2 years before the bankruptcy has the ability to predict a right thing. Therefore, the results have been shown that all four models were statistically significant. Consequently, there are no significant differences. All models have the precision to predict the financial crisis.