Dynamic pricing for perishable products using reinforcement learning

Revenue management is one of the successful applications of operations research. The area of revenue management we are concerned with, in this thesis is dynamic pricing. More specifically, we are concerned with establishing a pricing policy that maximises the revenue for a fixed number of products or services in a finite selling horizon. The characteristics of the industries considered are, the full inventory of items is available for sale at the beginning of the selling period, no re-ordering is allowed, and the items unsold by the deadline have a constant salvage value of zero