University of Twente Student Theses


Optimal Forgetting: Dynamic Pricing in Changing Markets

Heegde, L. ter (2015) Optimal Forgetting: Dynamic Pricing in Changing Markets.

[img] PDF
Abstract:In this thesis we consider dynamic pricing and learning in changing markets. When estimating the demand function, a certain amount of previous sales data is taken into account. In case of changing markets it might not be optimal to take all available sales data into account in the estimation. The goal of this research is to find the optimal amount of sales data ($N_t$) that we have to include in the demand estimation in every time period. Analytical results are given for a deterministic price set. A constant market is analysed as well as a market with a change point. We show that already in a constant market it is not as simple as we might think. The estimation of the intercept namely also depends on the points added in the estimation. For the change point model we show that it depends on the size of the change in the model parameters whether or not we improve the estimation of the slope parameter by adding pre-change data. Simulations are performed and the Controlled Variance Pricing policy is used in them. The performance of five possible subsequences for N_t is compared in five different scenarios for the model parameters. In a market with no or only one small change point it is optimal to take all available data into account. In markets with one large change point, or with continuous change, it is optimal to choose N_t small. With the described method we can improve the performance of existing pricing policies.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:31 mathematics
Programme:Applied Mathematics MSc (60348)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page