价格指数是衡量物价波动的重要指标,特别是居民消费价格指数和商品零售价格指数,近年来受到广泛关注。如能预测出价格指数的大致走势,将对决策者和居民都具有重大意义。与点预测和区间预测相比,密度预测包含的信息更多,得到的预测结果也更有价值,因而值得在价格指数预测中加以应用。将密度预测法与常用的时间序列模型相结合,并考虑预测误差可能存在的一阶马尔可夫性质,可实现较好的预测效果。 Price index is a measure of price volatility important indicators, especially the consumer price index and retail price index, in recent years by the widespread attention. Predicting the general trend of the price index will be of great significance to policymakers and residents. Density forecasting contains more information than point forecasting and interval forecasting, and the resulting forecasting result is more valuable and therefore worth applying in price index forecasting. The density prediction method is combined with the commonly used time series model, and the first-order Markov property which may exist in the prediction error is considered, and a better prediction effect can be achieved.