Analysis And Prediction, Based On Data Warehouse, Cotton And Oil Prices

This paper discusses the recent developments and the wide application of data warehouse technology in various fields at home and abroad, and studies the application of such technology to the price trend analytical system of grain, cotton and edible oil for the purpose of providing managerial decision-making support. In addition, the paper constructs a data warehouse platform oriented to analysis and decision-making support, and provides an analytical and forecasting method for analyzing and forecasting the price data of grain, cotton and edible oil by adopting the Markov and the feed-forward neural network analytical methods. These studies serve as complements to the inadequacies of the current analytical technology. These studies include the following aspects:(1) a discussion of data warehouse technology and on-line analytical technology, and their applications;a comprehensive analysis of the current methods;a demand analysis of the data warehouse technology-based priceanalytical system of grain, cotton and edible oil on the basis of data warehouse modeling;the tentative construction of a platform useful not only for regular inquiry and statistical applications but also for analytical and decision-making purposes;the modeling of the price data warehouse of grain, cotton, and edible oil by means of extracting, transforming and loading data.(2) a study of the Markov model and the provision of suggestions to the application of the Markov method with specific calculations.(3) a study of the structure of feed-forward neural network;a discussion on the regular construction method of network model based on time data series forecasting by feed-forward neural network;and a discussion on the actual forecasting of the price data of grain, cotton, and edible oil through BPnetwork.(4) a discussion of the causes of the problems in the forecasting, and the suggestions for solving the problems.