Research On Grey Target Decision Method And Its Application Based On Temporal And Spatial Attributes Of Panel Data

The theory of grey system was established in the 1980 s.After more than 30 years of development,it has formed a relatively perfect theoretical system,and the grey decision making is one of the important research branch of the grey system theory,Among them,the grey target decision-making method has received widespread attention since the day it was put forward.Through the tireless efforts of many scholars,it has been constantly developed and perfected,and has been successfully applied to many fields such as economy,management and social production practice.It has solved a lot of practical problems.In the existing research on the grey target decision method,most of the models are suitable for the multi index static decision problem with the cross section data as the information carrier.However,there are more and more multistage dynamic decision-making problems in life.How to solve the problem of more and more multi-stage uncertainty decision making,how to realize the effective mining of information in this kind of problem and construct the model and method system of solving this kind of problem has become an important issue to be studied urgently.Cross section data is to obtain data from different subjects at the same time;panel data is to intercept a number of cross section data in a sequence of time sequences.Panel data can include multiple indicators,multiple schemes,and multiple stages of information at the same time.In multistage decision making systems,more and more decision information is presented in the form of panel data.Based on this,this study uses the interval grey number panel data to represent the decision information with uncertain characteristics.The main content of this paper is: without considering the time value of information and under the condition of considering the time value of information,based on the structural characteristics of panel data,builds a grey target decision-making model which can deal with panel data decision information,in order to expand the system of grey target decision method,it lays the foundation for expanding the decision method system based on panel data.One is to build a grey target decision model based on the temporal and spatial attributes of panel data without considering the time value of information.The decision-making process of the model is as follows: first,draw lessons from the thought of the unit circle,construct circular-like graphics,the graph area is used to represent the performance of the sample(scheme)at the fixed time point,to realize the aggregation of single stage decision information,and the static target distance is obtained according to the principle of grey target decision.And then,construct the same index difference sequence of adjacent period,combined with the index weight characterization of samples(solution)development trend in a period of time,to realize the aggregation of decision information in time dimension,according to the principle of grey target decision,the dynamic target distance is obtained.Finally,the static target distance and the dynamic target distance are assembled by the function to obtain the comprehensive target distance,and complete the decision-making process.The model is applied to the case of selecting the best supplier,and the reasonable decision results are obtained and the feasibility of the model is verified.Second,in considering the time value of information,construct the grey target decision model based on the spatial and temporal attributes of panel data.Information has timeliness,specifically refers to with the passage of time,the value of information will be reduced gradually,so the introduction of the time value of information function,eliminate the effects of the information entropy,the qualitative expression of information time value function is composed of information stock,information utility and information loss.The transformation of three factors from the qualitative expression to the quantitative expression is realized by the method of analogy,thus the quantifiable time value function is obtained.The decision-making process of the model is as follows: Firstly,according to the third chapter,the single stage decision information is extracted to calculate the performance values of each plan at all time points.Then the time value coefficient of the information is calculated.Finally,the time value coefficient is brought into the formula of the target distance,to obtain the comprehensive range,and complete the decision-making process.Applying the model to the same case,the decision results are basically the same as those in the third chapters,which verify the feasibility of the model.This research provides a method to solve the multi-stage decision problem,and improves the grey target decision method,and the application scope of grey target decision method is extended from two-dimensional plane to three-dimensional space.