基于Markov链的最优化预测模型及其应用研究

A Research on Optimal Forecasting Model and Its Application Based on Markov Chain

马尔可夫预测方法在预测领域有着广泛的应用.该方法应用的一个重要的问题就是如何估计一步状态转移概率矩阵.在历史资料没有给出系统处于n个状态次数的情况下,给出一步状态转移概率矩阵估计的最优化方法.最后探讨了基于M arkov链的最优化预测模型在长江水质预测中的应用,从而表明该模型的有效性.

Markov forecasting method has wide applications in forecasting fields. A very important problem of applying this method is how to estimate one-step state transition probability matrix. When the amount of n states of the system by historical data is unknown, the optimal model is given to estimate it in this paper. In the end the optimal forecasting model is applied to forecast the water resource in Changjiang River based on Markov chain. The result shows that the model is effective.