Prediction to Industrial Added Value Based on Holt-Winters Model and ARIMA Model

摘要

Industrial added value (IAV) is an important indicator to measure a country's industrial development level, at the same time, is also a core indicator of national economy accounting system. The paper respectively using ARIMA and Holt - Winters, the two time series forecasting model, to fit the monthly data of industrial added value of Hubei province in 2009-2014 and on the basis of it we forecast for next year. For ARIMA model, this paper introduces the smooth sequence processing, model identification, parameter estimation and model diagnosis and prediction, such as the key to predict; For Holt-Winters model, this paper introduces the selection of smoothing coefficient alpha, beta, gamma and initial smoothing factor, recursive calculation process, such as the key to predict, and the corresponding algorithm and the prediction model is designed. Finally use example to compare the two methods and analyze the pros and cons of two kinds of models to predict.