The accuracy of sales forecast has great impact on manufacturing and sales. With the quick development of the society, it is really hard for traditional forecast system to meet new demand in dealing very large amount of data and sales forecasting with high accuracy. Obviously, data mining technology is the key to solve those problems. In this article, the first two parts are a brief introduction about the concept of Sales Forecast, then an introduction of two popular Extract-Transform-Load (ETL) tools and a comprehensive analysis of four most often used forecast method are given in the following parts. Finally, based on analysis and combined with a variety of technical advantages, we propose a new sales forecast system model based on data mining and Grey-Markov prediction model is used as an example to illustrate its working principle and to verify its feasibility theoretically.
展开▼
