|
Predicted There are many , but most of the methods are based on the case of a large number of samples can be implemented or were able to obtain higher accuracy , but we are not in any case be able to collect the complete sample data , or the study of a some object is very limited due to the nature of the research questions selected sample size , gray system prediction is for the poor , the small sample size of the object to predict research . Markov model combines gray theory to avoid considering the impact of a variety of other factors , the state in which object may also be predicted . In this paper, starting from the basic theory of gray model and Markov chain for traditional GM (1,1) model is deviation the gray exponential model , the model 's accuracy is not high , the proposed residuals GM (1,1 ) and metabolic GM (1,1), Markov transition probability combined reflect the inherent regularity advantages of transfer between the impact of random factors and state , to make full use of the information of the original data , and solve the gray prediction model for random volatile series low prediction accuracy , a combination of both to avoid considering a variety of other influencing factors , they can predict the future time changes , with greater scientific and practical . To test the validity of the prediction model , the total import and export volume of the country and the ASEAN gray modeling , metabolic GM (1,1) and Grey Markov , the total import and export volume of import and export volume between Yunnan and ASEAN and Guangxi and ASEAN gray modeling , residual GM (1,1) Grey Markov model results contrast gray Markov model prediction accuracy than gray prediction model has high practical value .
|