Σύγκριση υποδείγματος Logit και μοντέλου Τεχνητών Νευρωνικών Δικτύων για την πρόβλεψη εταιρικής αποτυχίας για δείγμα εταιρειών NYSE [article]

Nikolaos Karaliotas, National Technological University Of Athens, National Technological University Of Athens
2015
This thesis begins with the first chapter, where there is a reference on the definition of corporate failure. In the beginning there is an introduction regarding this definition according to several scientists and researchers on Economic Analysis and corporate failure prediction models. A distinction is being made between corporate failure and bankruptcy (formal default), so as to avoid any misinterpretation between the two terms. At the end of the chapter there is a synopsis and final argument
more » ... regarding the corporate failure term. The second chapter consists of a literature review regarding the corporate failure prediction models and it is separated in two parts. The first part is a historical view and presentation of the evolution of scientific work on the prediction of corporate failure. In the second part, there is a thorough analysis about the statistical univariate methods of corporate sorting according to financial health, and to be more specific, there is a review on Univariate Analysis, Multiple Discriminant Analysis -MDA, the Linear Probability Model -LPM, the LOGIT and the PROBIT model. The third part of the chapter sees a detailed analysis on the modern techniques of corporate failure prediction, which actually consists of Neural Networks -NN, Support Vector Machines -SVM, Evolutionary Algorithms -EA, and several other methods. At the end there is a commentary review regarding all the models of corporate failure prediction.
doi:10.26240/heal.ntua.4494 fatcat:upgmszzgfzflbied3ajum5naz4