A discrete teaching-learning-based optimisation approach for e-commerce product image placing and inventory planning

Abstract

The visual-attention-dependent demand (VADD) model with genetic algorithms has been proposed for the product placement and inventory management problem of e-commerce websites. In the face of an e-commerce environment with continual changes in potential products, customer needs and competitors, this study proposes a method that modifies traditional continuous TLBO algorithm with a discrete-continuous conversion technique to solve the VADD model. To verify the usability and efficiency of the proposed method for the VADD model, this research compares the proposed method with the exhaustive search and the genetic algorithm (GA) for large-, medium-, and small-scale problems. In order to enhance the efficiency of methods, the values of their control parameters are set by Taguchi method, respectively. Results show that both the GA and the proposed methods have excellent performance in the approximation ratio, but the proposed method spends less time than the GA method, particularly for large-scale problems. Thus, the proposed method could help e-commerce merchants rapidly making decisions on the product placement and inventory management to increase the sales of goods.