![[2602.16842] What is the Value of Censored Data? An Exact Analysis for the Data-driven Newsvendor](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.16842] What is the Value of Censored Data? An Exact Analysis for the Data-driven Newsvendor
Summary
This paper analyzes the impact of censored demand data on inventory management, presenting a method to compute worst-case regret for data-driven policies and highlighting the importance of accurate point-of-sale information.
Why It Matters
Understanding the value of censored data is crucial for businesses relying on inventory management. This research provides insights into optimizing inventory policies under conditions where demand is not fully observable, which can significantly affect profitability and operational efficiency.
Key Takeaways
- Censored demand data limits learning from sales data, affecting inventory decisions.
- A new method reduces complex optimization problems to finite dimensions for better policy evaluation.
- Targeted exploration at high inventory levels can enhance performance under heavy censoring.
- Sales-as-demand heuristics can lead to significant performance degradation when data is censored.
- Accurate point-of-sale systems are critical for effective inventory management.
Computer Science > Machine Learning arXiv:2602.16842 (cs) [Submitted on 18 Feb 2026] Title:What is the Value of Censored Data? An Exact Analysis for the Data-driven Newsvendor Authors:Rachitesh Kumar, Omar Mouchtaki View a PDF of the paper titled What is the Value of Censored Data? An Exact Analysis for the Data-driven Newsvendor, by Rachitesh Kumar and 1 other authors View PDF Abstract:We study the offline data-driven newsvendor problem with censored demand data. In contrast to prior works where demand is fully observed, we consider the setting where demand is censored at the inventory level and only sales are observed; sales match demand when there is sufficient inventory, and equal the available inventory otherwise. We provide a general procedure to compute the exact worst-case regret of classical data-driven inventory policies, evaluated over all demand distributions. Our main technical result shows that this infinite-dimensional, non-convex optimization problem can be reduced to a finite-dimensional one, enabling an exact characterization of the performance of policies for any sample size and censoring levels. We leverage this reduction to derive sharp insights on the achievable performance of standard inventory policies under demand censoring. In particular, our analysis of the Kaplan-Meier policy shows that while demand censoring fundamentally limits what can be learned from passive sales data, just a small amount of targeted exploration at high inventory levels can...
