Study On Uncertain Multi-attribute Decision Making Algorithms Based On Set Pair Analysis

The uncertain multi-attribute decision-making has become increasingly importantin people’s real life, because of the complexity, uncertainty of the objective things andambiguity of the human mind. Use of these theories deal with uncertainty inmulti-attribute decision-making is being concern and develops rapidly with some newtheory and methods which dealt with uncertain information are proposed.This article combined Set Pair Analysis (SPA) theory with the uncertainmulti-attribute decision making, study in uncertain multi-attribute decision makingalgorithms based on the Set Pair Analysis. Set Pair Analysis theory is a newmathematical tool for analyzing fuzzy and uncertain knowledge; it is more capable thantraditional theories of uncertainty in describing uncertain information. The currentresearch of Set Pair Analysis is not yet mature. This paper studied the attribute valuestandardization, attribute weights, and multi-attribute decision algorithm based on SetPair Analysis, and presents new insights. The main contents are as follows:①For the evaluation of multi-attribute decision-making system, considering theclassification standard threshold values are fuzzy under the circumstances that theclassification standard of index is known, the value of the property are normalized byusing Contact degree of Set Pair Analysis, and studied a method for subjective weightand objective weight calculation by computing of Contact degrees, proposed theevaluating algorithm based on Analytical Hierarchy Process and Set Pair Analysis(AHP-SPA),and used contact number of Set Pair Analysis express the results.②For the sort of multi-attribute decision-making system, under the circumstancesthat the classification standard of index is unknown, consider that the more difficult themore property values to the right direction, study the non-linear representation ofcontact degree of attribute value in comparative zone, and the study of a new IdentityDifference Opposition (IDO) distance function, proposed an TOPSIS sorting algorithmbased on weighted IDO close degree, sorted by IDO close degree size.③For the prediction of multi-attribute decision-making system, taking into accountthe multi-attribute problem is difficult to be predicted by modeling the physical cause,proposed a prediction algorithm based on time series weighted SPA-Markov Chain.Express the state of Markov chain by contacting degrees, and then calculate statetransition probability matrix of the adjacent time series based on the Monte Carlo method, and assigned the Markov chain weights through the similarity degree of IDO,at last, departure from Markov assumptions, calculations the predict results next time.The paper analyses the feasibility of the algorithm by using urban ecologicalsecurity assessment which is a multiple attribute decision-making problems, the resultsprovide some reference value and scientific basis for the decision-making.