点击次数：

论文类型：会议论文

发表时间：2016-11-04

收录刊物：EI、CPCI-S

卷号：660

页面范围：115-129

关键字：On-line products; Diffusion; Symbolic regression; Seasonality; Bass model

摘要：With the popularity of on-line shopping increasing, on-line products diffusion becomes the growth of importance for marketing decision. Bass diffusion theory is a classical method of forecasting products sales. In this paper, we introduce symbolic regression method to describe the trend of on-line products diffusion and verify whether Bass model in on-line condition performs good as before. Almost all products exhibit seasonality in their sales pattern. Considering the particularity of on-line shopping, we define monthly, weekly and daily period as the division of on-line season. The models perform differently when frequency of data varies.