THE İNVESTİGATİON OF ARTIFICIAL NEURAL NETWORKS AND DISCRIMINANT ANALYSIS IN PREDICTION OF FINANCIAL FAİLURE AFTER GLOBAL CRİSİS
KÜRESEL KRİZ DÖNEMİ SONRASI FİNANSAL BAŞARISIZLIĞIN DİSKRİMİNANT ANALİZİ VE YAPAY SİNİR AĞLARI İLE İNCELENMESİ

Author : Tuğba GÖKDEMİR -- Sezai GÖKDEMİR
Number of pages : 369-387

Abstract

The failure of a commercial company is a significant loss to its buyers and shareholders For this reason, establishing models that can predict foreseeable failures provides great convenience and warns those who are injured. In this study, it is aimed to determine the financial ratios affecting the financial failure of the manufacturing industry firms traded on the Istanbul Stock Exchange and to predict the financial failure. For this purpose, artificial neural networks and discriminant analysis were used. Using the data between the years 2008-2013 of the manufacturing industry firms which traded in ISE, the three dependent variables models which different from similar studies in the literature next to the model, which includes all failure criterion failure discussed by the debts exceed assets criteria and overlap located between two or more years of damage to have been developed and classification success of disciriminant analysis were compared for this model. In there sult of the study, it was determined that the method that best classifies financial failure is artificial neural networks. In addition, when the financial failure is determined, it is determined that the criteria of exceeding the equity of the debts is a more effective criterion for financial failure and the most important financial ratios determining the financial failure are the profitability and financial structure ratios.

Keywords

Financial Failure, Disciriminant Analysis, Artificial Neural Networks

Read: 762

Download: 244