NUSRI-CQ PhD student Liu Jingren participated in the best paper competition and won the third place at the 4th National Conference on Supply Chain and Operation Management in Shanghai on December 4, 2023.

The paper's title is "Pricing Analytics with Shape-Restricted Demands", co-authored with Qin Hanzhang, Assistant Professor, Industrial Systems Engineering and Management, NUS, and Mabel C. Chou, Associate Professor, Operations and Analysis, NUS, and Senior Principal Investigator, Modern Logistics Centre, NUSRI-CQ.
 

Prof Qin Hanzhang (left) and Liu Jingren (right)
 

Abstract:
We consider a fundamental problem in revenue management: feature-based pricing, where a firm needs to make a pricing decision to maximize the expected revenue for a single product based on feature information (e.g., weather, holidays, product attributes, etc.). Historical data with price, covariates, and uncensored sales, are available for demand estimation. Our model assumes a linear relationship between price and demand while simultaneously capturing the impact of covariates through a nonparametric shape-restricted function. We develop a Three-Step Semi-Parametric Estimation algorithm to estimate the demand and foster near-optimal data-driven pricing decisions. From a non-asymptotic perspective, we derive finite sample regret bounds, showcasing the efficacy of our algorithm in achieving near-optimal revenue, even under potential misspecification of the demand model. The numerical results demonstrate that the decision performance of our algorithm is comparable to the Double Machine Learning method, while significantly outperforming a naive two-step iterative learning method as well.

Background: In reality, customer demand is influenced by many factors, including price and features (e.g. weather, holidays, product attributes, etc.). Given the increasing accessibility of historical sales data, it is worth investigating how firms should learn the demand model and make pricing decisions based on the data.

Research methodology:
This study proposes a data-driven pricing policy to help firms make pricing decisions based on historical data. In demand estimation, our model assumes a linear relationship between price and demand while simultaneously capturing the impact of features through a nonparametric shape-restricted function.

Conclusion:
The shape-restricted demand model preserves good explainability and computational efficiency, which supports pricing decisions aimed at maximizing revenue. It is worth noting that although the shape-restricted demand model might suffer from misspecification issues, our theoretical and numerical results demonstrate that the pricing policy based on such a misspecified model can generate near-optimal revenue.