Artificial intelligence techniques for modeling financial analysis

Abstract

Although monitoring financial health of small firms is decisive to their success, these firms commonly present difficulty when analysing their operational financial condition. In order to overcome this fact, the present thesis proposes a financial knowledge representation that is capable of proposing alternative actions whenever a deviation is detected. The knowledge representation developed recognizes the existence of different phases of analysis: one that looks for some clues about possible financial problems and another one that focuses on with more detail the potential problems detected by the prior phase.The vagueness present in many semantic rules was implemented by using the Theory of Fuzzy Sets. The uncertainty about the future behavior of some key financial variables is incorporated by means of managers perceptions about trends and events. A practical formulation of this proposal is done considering the retail bus sector.