Grey Trend Relational Analysis And The Application

The trend forecasting is the recognition and the forecasting which occur to the each kind of information of the things in the developing process,according to historical condition and current condition to speculate future development tendency. Until now,the research technique roughly may divide into three kinds:The statistics type,the successive type,the recursion type.Because the method of forecasting and statistics should determine the statistical data conform to what kind of probability distribution,therefore the space of the sample should be a litter big,and it is easy to make big error;The method of the fuzzy forecasting is based on the artificial intelligence and this method takes the knowledge engineering as the guide to forecast and fully uses the analysis ability of the humanity to gain,pass,process,regenerate and use information effectively,at present it is in the stage of the exploration;the modern forecasting methods,for example the grey system forecasting method,are developing on the succession model,at present they are moving towards the practical stage from the research stage.This article has first summarized the grey system theory,as well as the tendency forecasting method,and simply introduced several kinds of grey relation which are commonly used,and separately indicated their good and bad points.Then described the similar systems engineering theory,and profited from the thought of this theory to infer the characteristic that the trend relations of the same thing between continual two time intervals are nearly equal,and establish the forecasting equation based on the above characteristic.In the computation process,because the trend relational formula which is quoted in this article has the absolute value,so it should have the positive and negative this two kind of situations,therefore uses orthogonal table technology which has the extremely widespread use to calculate the data,simplified the process of the operation;then carries on the solution by using MATLAB.Finally through the analysis of the computation of the example,the result indicates the precision of this model is higher than it of the GM(1,1).Afterwards,on the foundation of an original trend relation,the two-step trend relation is proposed in this article,and we hope carry on the revision to the original trend relation formula through it.After analyzing the conclusion of the example application,it could be obtained that the result of trend relational analysis though the two-step trend relation is better.At the end,this article applies the grey relational analysis in the analysis of the random sample,it has been separated from the former traditional statistics analysis way,causes the sample to be allowed to use "few data" to analyze the population,and infers the relational formula;and also applies the grey relational analysis in the present rising subject—complex network,through analyzing the three important factors—the clustering coefficient,the average path length and the degree,do relational analysis to the two complex networks which have the same point number or nearly same points,and the degree distribution of one network is power law distribution and it of the other network is the exponential distribution.