Title :
Dynamic pricing and reinforcement learning
Author :
Carvalho, Alexandre X. ; Puterman, Martin L.
Author_Institution :
British Columbia Univ., Vancouver, BC, Canada
Abstract :
We consider the problem of optimizing sales revenues based on a parametric model in which the parameters are unknown. The manager has to set the price at a level in order to maximize current revenues and at the same time learn about the parameter values to increase the future revenues. Both demand and price are assumed to be continuous variables. We study several different strategies for learning and show that a one-step look-ahead rule produces good short term performance.
Keywords :
learning (artificial intelligence); optimisation; parameter estimation; pricing; sales management; continuous variables; dynamic pricing; one-step look-ahead rule; parametric model; reinforcement learning; sales revenue optimization; Bayesian methods; Business; Infinite horizon; Learning; Marketing and sales; Noise level; Parametric statistics; Pricing; Stochastic processes;
Conference_Titel :
Neural Networks, 2003. Proceedings of the International Joint Conference on
Print_ISBN :
0-7803-7898-9
DOI :
10.1109/IJCNN.2003.1224034
