Harnessing Dimensionality Reduction with Neutrosophic Net-RBF Neural Networks for Financial Distress Prediction

Abstract

Neutrosophy is the study of neutralities and extends the discussion of the truth of opinions. Neutrosophic logic may be employed in any domain, for providing the solution for the ambiguity problems. Several real-time data experience problems such as indeterminacy, incompleteness, and inconsistency. A fuzzy set provides an uncertain solution, and intuitionistic fuzzy set handles incomplete data, but both fail to manage uncertain data. Before bankruptcy, financial distress is the early stage. Bankruptcies caused by financial problems can be seen in the financial statement of the company. The capability to predict financial problems became a crucial area of research since it provides earlier warning for the company. Moreover, predicting financial problems is advantageous for creditors and investors. In this article, we develop a new Dimensionality Reduction with Neutrosophic Net-RBF Neural Networks (DR-NSRBFNN) technique for FCP process. The DR-NSRBFNN technique concentrates on the predictive modelling of financial distress. In the DR-NSRBFNN technique, two major stages are involved. In the preliminary phase, the high dimensionality features can be reduced by the use of arithmetic optimization algorithm (AOA). In the second phase, the DR-NSRBFNN technique applies the NSRBFNN model to predict financial distress. The performance evaluation of the DR-NSRBFNN technique can be examined using distinct aspects. The widespread study stated the improved performance of the DR-NSRBFNN technique compared to other systems