Foreground Construction Method Based On Digital Probability And Text Probability

In the current rapid development of the information age, due to extensive participation in group decision making, the uncertainty related to decision making problems is increasing and decision problem becomes more complicated. In decision analysis, Prospect Theory has a good explanation ability to the people’s non rational decision-making behavior in the uncertain context and the decision maker’s risk preference, so it is favored. However, the decision analysis based on prospect theory of existing literature has a basic hypothesis, that outcomes and probabilities are known (structured data). But in fact, what decision information people can obtain is often a lot of uncertainty, unstructured information form, and that has posed a challenge to the application of prospect theory to solve the uncertain and complex decision problem.In this paper, using the method of information fusion, prospect building method is proposed based on evidence theory. According to that evidence theory can deal with the uncertainty brought by unknown and fuse multi-source using information combination rules, then get a more comprehensive, more accurate information for decision making. In addition, because of limited rational of human thinking the fuzzy understanding to problems, put forward a prospect building method based on the verbal probability. The main works are as follows:(1)For prospect building problem based on Dempster-Shafer Theory, on the basis of number probability, combined with the evidence theory, we proposed a prospect building method. The concluded inference information easily to be incomplete is considered in complex environment, and the alternative outcomes are grouped by experts to make sure the outcomes in one set with the same occurrence possibilities based on the set of all alternative’s outcomes. After that, the relative comparison judgment method and the complementary judgment assignment thought are combined to construct the probability inference matrix to reflect the subjective possibility cognition of each outcome group, and a transformation theorem is introduced to derive Basic Probability Assignment (BPA) from probability inference matrix. At last, each expert is considered as an independent evidence source, then use the Dempster combination rule and the DSmP transformation method to introduce prospect building procedures.(2) For inferring and building alternative prospect problems by the probability form of verbal probability, the basic analysis idea was as follows. Alternative outcomes possible set (frame of discernment)were grouped by experts, then experts made inference matrix in the form verbal probability according to relative comparison judgment, and through the conversion rules between the number and verbal probability, converted verbal probability matrix to probability judgment matrix, consisting of triangular fuzzy number. After that, constructed the optimal Basic Probability Assignment (BPA) function determining model to solve optimal BPA function. Used the Dempster combination rule and the DSmP transformation method to fuse BPA functions from all experts and to transform the probabilities. So, achieved the alternative’s prospect building based on verbal probability.To sales prospect estimation of the Hengda Real estate project as an example, this paper applied this method to construct prospects of the Royal Scenic project and Garden Scenic project. For Royal Scenic project, according to the expert’s reasoning ability, experts give uncertain inference information with numerical probability; for the Garden Scenic project, experts give uncertain inference information with verbal probability. Then estimated two projects’ sales prospects to provide important reference value for buyers, investors, construction companies, materials suppliers and other related enterprises.