关键词: 订单量, 百度指数, 组合预测

Abstract: Under the explosive growth of e-commerce orders, the ability to apply order-derived data is highly demanded. How to utilize the data for fast and dynamic prediction is the key point of online-shopping behavior study. In order to improve the stability of prediction model, this paper proposes a combination forecasting model based on integration of web search index, which incorporates BP neural network, Adaboost-based BP neural network and Support Vector Machine（SVM）. The paper also constructs an index system integrating Baidu index and order-derived information, for the sake of enhancing accuracy. Final contrast experiment results show the effectiveness of using web search indexes as an influence factor of forecasting model.

Key words: order, Baidu index, combination forecasting