The Use of Fuzzy, Neural Network, and Adaptive Neuro-Fuzzy Inference System (ANFIS) to Rank Financial Information Transparency

Abstract

Ranking of a company's financial information is one of the most important tools for identifying strengths and weaknesses and identifying opportunities and threats outside the company. In this study, it is attempted to examine the financial statements of companies to rank and explain the transparency of financial information of 198 companies during 2009-2017 using artificial intelligence and neural, fuzzy and neural-fuzzy network models. Accordingly, the best method to rank financial information transparency is selected. For this purpose, the information about companies in different industries is first sorted using the corporate financial statements in Excel software and then, the ranking of companies in each industry is determined on a scale of 1 to 5 in terms of financial and technical strength in the form of a diagram. In order to rank companies with artificial intelligence, the information obtained has been entered into Matlab software and neural, fuzzy and neural-fuzzy models are then implemented. After reviewing descriptive statistics and Fisher's test, companies are ranked. According to the results of the research, the best method for ranking is the neural method and the neural-fuzzy method. The results of the neuro-fuzzy method with 0.01 distance from the results of the neural method provide the best results after the results of the neural method. But in the fuzzy method, the ranking is far from the intended results and is not suitable for ranking of financial information.