A Dynamic Pricing Method in E-Commerce Based on PSO-trained Neural Network
Recently, dynamic pricing has been a common competitive maneuver in e-commerce. In many industries, firms adjust the product price dynamically by the current product inventory and the future demand distribution. In this paper, we used particle swarm optimization (PSO) algorithm to train neural networks, then introduced the PSO-trained neural network into e-commerce and presented a new dynamic pricing method based on PSO-trained neural networks. In the method, from production function principles we obtained the least variable cost, and by making the error of mean square between the actual outputs and expectation outputs minimal we got the optimal dynamic price of products. The PSO-trained neural network can simplify the rapid change of prices and can successfully set the optimal dynamic prices in e-commerce.
KeywordsParticle Swarm Optimization Marginal Cost Particle Swarm Optimization Algorithm Selling Price Dynamic Price
- 5.X. J. Wang, Y. L. Chen and W. Q. Zhu, The Pricing Strategies for Agents in Real E-Commerce, WINE 2005, LNCS 3828,887–894(2005).Google Scholar
- 8.L.C. Jiao, Neural Network System Theory, Xidian University Press, Xi an(1996).Google Scholar
- 9.S.A. Cong, Face the MATLAB toolbox nerve network theory and the application, University of Science and Technology of China Press, Heifei(1998).Google Scholar