A fuzzy hybrid system for supporting financial analysis

Paper presented at the Third Congress of the International Association for Fuzzy-Set Management and Economics, Buenos Aires, Argentina.The correct perception of the financial condition of a firm is dependent upon the quality of its information processing system and upon the analyst’s expertise. The more complex the financial environment, the more difficult is the identification of problematic situations. Monitoring and correcting inappropriate conditions become critical tasks, particularly to small and medium companies, where the presence of an expert is not affordable. For such companies, it is of great value to have a system available capable of spotting the condition and suggesting alternative actions to control anomalies. We have developed a hybrid intelligent system with such features. It is composed of two modules: the first one identifies the financial condition and the second one suggests possible actions to be taken in order to revert the deviations. Particularly in the second module, we have identified two facts in the knowledge acquisition process: first, solving a financial problem, once its type is known, is a deductive task, in the sense that it is accomplished by comparing facts with associated patterns. Second, linguistic terms such as “high reduction of inventory” and “low interest” were constantly used to describe actions to be taken or information about the economy in a certain time. These two features lead to the development of a fuzzy expert system capable of indicating operational actions to control problematic financial situations. In this article we detail the fuzzy financial advising system module