Study of rainfall prediction model based on GM (1, 1) - Markov chain

This article adopts the method of Gray Markov to predict the rainfall. Gray GM (1, 1) model is used to establish the rainfall prediction model with the gray system composed of rainfall over the years. It is poor fit for random and volatile data sequence; therefore, the prediction accuracy is also low. However, the Markov chain can describe random change and dynamic system. It mainly based on the transition probability between the different states of the subjects to infer the systems´ future development. Because the problem about the prediction of rainfall changes over time and shows a trend of non-stationary stochastic process. And it is subject to various random factors. Therefore, combine Markov prediction model with the gray prediction model necessarily. By using their advantages, greatly improve prediction accuracy of the random and volatile data. So it can provide a new way to predict the Volatile random objects.