Design And Realization Of Fuzzy Decision Making Data Analysis Platform Based On Fuzzy Regression

Fuzzy decision making is outcome of combination of fuzzy sets and decision theory,which has offered complete theory basis and method models to solve half structureddecision making problems and not structured decision making problems. This paper firststudies the current some kinds of classic fuzzy decision making methods, and comparestheir suitable condition and application effect with contrast and analysis , puts forward someshortcoming and shortage in application,which mainly expressed in decision makingcourse , person participating in too much and method lacking flexibility. Furthermore thesemethods look so inextricable to those decision making problems whose scheme is nearlyimpossible to be put forward. According to these problems, this paper then puts forward the concepts of fuzzyexpectation,fuzzy variance and fuzzy least-square method, moreover takes these conceptsas theory basis to put forward fuzzy decision making model based on the analysis ofregression, which can carry out many regression calculation, can make the bad informationand the data of morbid state in measure,observation and actual information for decisionresult disturb to reach minimum, can get the result that we are compare satisfied with. Aimportant characteristic of this kind of model is, it can get rid of the intervention of personand remove the uncertainty that artificial factor brings in the decision making course. Final, this paper establishes a data analysis platform which has taken the fuzzyregression model as starting point. This platform can carry out the entire course ofregression analysis, which has data record into function,data management function ,graphto present function ,system installation and corresponding help function. The platform hasbut development, other fuzzy decision making models based on mathematical statistics andanalysis can join it continuously, which make this data analysis platform function perfectand method complete in order to satisfy different demands of user's . The platform hastransplantability, which offers general interface to other use. We can make the modelreceive other application platform relaxedly, it has offered maximum convenience to solvesimilar problem.