Font Size: a A A

Research Of Short-term Sale Forecasting Methods

Posted on:2005-04-01Degree:MasterType:Thesis
Country:ChinaCandidate:L N WangFull Text:PDF
GTID:2156360122985411Subject:Systems analysis and integration
Abstract/Summary:PDF Full Text Request
The economic prediction makes policies services for economy. The goal is to improve the science level of economy management, and reduce the blindness of economy decision. To decrease indeterminacy of future and venture of making economy decision, it is important to hold the economic development and understanding the change trends of the future market .Along with the development of domestic economy and the improvement of purchasing power, the market demand is increasing in number. Considering making full use of feedback of market information to meet the requirement of business activity, market forecasting must be various and frequent. Market forecasting ,to the end , is to predict the requirement of market(it says sale for the enterprises ,too).Short-tenn sale forecasting reflects the facts of market change in good time .It is significant to study short-term sale for market forecasting.Short-term sale forecasting summarizes the law of commodity sale from the early sale, and predicts the future sale automatically. The data increases severely in which a lot of targets in fact are so complicated and uncertain to modeling or very difficult accurate modeling, and this proposes the new direction of the research of short-term sale forecasting technology.This paper thoroughly and deeply studies the theories and methods about short-term sale forecasting. After synthetically introducing the current situation of the common-used forecasting theory , and explains in detail the research of the ANN forecasting theory and its applied situation. Then proceeding from request for short-term sale forecasting ,the paper proposes time series model, fray prediction and BP learning algorithm ,and predicts the sale project in possible method .The paper adopts fray prediction by correlative factors ,making full use of the merits that fray prediction need a few datum and simple operation ,not considering the distributing rule and the change trend . And also adopts time series models by trend forecasting,of which the parameter evaluation is mature and the prediction is precise . Directed at the powerful nonlinear mapping ability of neural network .the paper sets up 3-BP network .On this basis, the paper sets up tentatively an combined forecasting methods based on the ANN theory which may accord with the short-term sale forecasting ,and offers the detailed theory method appears and design-steps. In order to observe the validity of method that the paper puts forward, the paper combines traditional forecasting methods and ANN theory , and the methods are applied to the retail sale project of the domestic society consumable monthly.
Keywords/Search Tags:ARMA model, Gray model, Neural Network, Combination forecasting
PDF Full Text Request
Related items