LinearPredictor

  一个线性预测的matlab仿真程序，先用模型生成原始信号(平稳随机过程)，然后利用最小均方误差准则计算出预测系数，最后利用预测系数对原始信号进行预测(利用原始信号过去n个时刻的值预测当前时刻的值)。(This is a matlab program for linear prediction. First, we use all-pole model to generate a signal(stationary signal), then utilize the Minimum Mean-Square Error to compute the filter cofficient, at last we can use the filter cofficient to predict(use past n signal data to predict current signal data).)