Research of Markov Chain to Predict the Commodity Retail Price Index Based on Entropy Weight

This paper aims to put forward a prediction analysis method of Markov Chain based on Entropy Weight, according to the changing characteristics of the commodity retail price index. Firstly, it applies equipartition method of the probability to partition the state grades; Secondly, it figures out the state of commodity retail price index, and constructs the state transtion probability matrix of the Markov Chain with differen step lengths; Finally, it not only predicts the comodity retail price index by using the Markov Chain analytical model based on Entropy Weight, but also uses an example of Harbin, from 1954 to 2008, to illustrate the validity of the method.