Research On Properties Concepts Based Financial Intelligence Decision-making Model

Financial market is a dynamic and complex system. The form of its price sequence is inherent complexly, not steady, and disordered, therefore, the decision- making models of the financial market are varied and far from mature.With the development of artificial intelligence techniques, especially the uncertainty theory of artificial intelligence, uncertainty denotation and reasoning mechanism were introduced into the financial decision-making model to establish and evolve decision-making rules. This technology can overcome the difficulties of building specific mathematical factors model that brought by the traditional methods for the financial market, which is complex and adaptive. It also brings a wide range of applications into the financial market with artificial intelligence techniques. In this paper, based on complex decision-making system modeling theory, a method of modeling information properties concepts and process mechanism was built; based on this, a new financial decision- making model, which used uncertainty knowledge transforming model-- cloud model; Bayesian probabilistic reasoning network; and evidence theory of information integration, was built. This paper mainly includes the following three aspects: based on properties concepts modeling and process mechanism; financial intelligence decision-making model; and the implementation of this model. The Properties Concepts based Financial Intelligence Decision- making Model based on (PC-FIDM) was described in detail. The main research contents are as follows:(1) Based on abstract"D_N"theory, we came up with the properties concept modeling and process mechanism which was built by uncertainty denotation and reasoning mechanism, its implementation was also discussed.(2) The frame of the financial decision-making model was established by combining the information process and properties concept modeling, and the data processing of this model was introduced in detail.(3) On the basis of Bayesian Probabilistic Reasoning Network, the decision– making prototype was implemented. After analyzing and dealing with the existing financial data, the efficiency of this model was validated.Compared with the traditional financial model, we designed a new financial intelligence decision-making model on the basis of information simplifying and properties concepts refining. The advantages are as follows: A more reasonable decision-making model frame denoted by uncertainty knowledge was established with the help of properties concepts modeling and process mechanism. The historic probability cases constructed by Bayesian probabilistic reasoning theory showed the feature of uncertainty of complex decision-making system. By using evidence theory, we expanded the methods of integrating information credibility. After analyzing and dealing with the financial data, this model can get several credible rules, which can be expanded and extended easily.