指数平滑法是对时间序列进行预测常用的方法之一,其理论基础是趋势外延，即未来一定时期内事物在数量上的演变特征不脱离过去的发展趋势,预测值是以前所有观测值的加权和,且对不同的数据给予不同的权数,越近期数据给予越大的权数,越远期的数据给予越小的权数。-Exponential smoothing method is commonly used to predict the time series one of the ways, the theoretical basis of the trend of the extension, that is, a certain period of time what the next evolution in the number of characteristics are not out of the development trend of the past to predict the value of all previous observations is the weighted and, on different data and different weights given to more recent data to give greater weights, the more long-term data to give the smaller weights.