Research Of The Forecast Of Apparel Sales Based On The Combined Forecasting Model

The sales rule of clothing merchandise is not only a simple linear pattern,it is also affected by comprehensive factors such as market,economy,policy,etc.,which contains a great deal of uncertainty and fuzziness,and contains inevitability to some extent.With the limited sales data,it is of great significance to choose reasonable predicting methods,to arrange the purchase-salestock of enterprises,and to reduce the operating cost.This paper first introduces the inventory backlog of these enterprises,instability between supply and demand,and the necessity of accuracy of sales prediction,and also introduces the research status in apparel sales prediction.For the linear,nonlinear,certainty and uncertainty problems in clothing commodity sales,the Exponential smoothing predictive model,moving average method predict model and the theory of Grey predict model of three single forecasting model are selected to study.And next,do the analysis comparison of the model of single instance.To determine the Grey theory and index smoothing predict model are combined to improve the predict accuracy through the analysis of the relative error of each single prediction.In order to solve the weights of the Grey theory-Exponential combination model,three methods of arithmetic mean combination predict,variance reciprocal method and entropy weight method were selected for comparison and analysis.In order to solve the problems that the votile will reduce influenced by external factors,on the basis of Grey theory model selection of Markov model for optimization,the time series model and dynamic model combined with economics model.Finally,there are specific prediction processes of the two combination prediction models,and the prediction accuracy of the two combined models proposed in this paper is tested by using the existing commodity sales data.The experimental results show that Grey theory and Markov model prediction accuracy is higher than grey theory-index portfolio model.Grey theory and Markov model is not only reduce the risk of single model's errors,but also solve the prediction accuracy which is influenced by external factors when the volatile will reduce the problem.