Sales Forecast of Dual-Channel Supply Chain based on Improved Bass and SVM Model

Abstract

As the increasingly fierce competition, more and more manufacturers ha ve started to sell products through online channels apart from traditional retail channel. Meanwhile, increased competition h as significantly shortened the product life cycle, which makes sales forecast meaningful. Taking lack of historical record into accou nt, the sales forecast model based on historical sales data based on Bass model as well as the improved Bass model by consideration of service and purchase factors are firstly discussed. And then, based on analysis of above model, a multi-factors sales forecast model based on SVM (Support Vector Machine) for short-life cycle product in dual-channel is creatively proposed, which integrates multi-factors, s uch as the improved Bass model predictive value, seasonal index, number of customers, credit value in practice. Finally, through simulation analysis, it is found that SVM-based multi-factor model is able to improve sales forecast of short life-cycle products in dual-cha nnel supply chain system.