Deep Neural Newsvendor

TL;DR

This paper introduces a deep neural network approach for data-driven newsvendor problems, providing theoretical guarantees and demonstrating superior performance on real-world and simulated datasets.

Contribution

The paper develops a DNN-based framework for quantile estimation in newsvendor problems, with non-asymptotic risk bounds and practical insights for real-world applications.

Findings

DNN method outperforms alternatives across various cost parameters

Theoretical excess risk bounds depend on smoothness and feature dimension

DNN performs well with both large and limited sample sizes

Abstract

We consider a data-driven newsvendor problem, where one has access to past demand data and the associated feature information. We solve the problem by estimating the target quantile function using a deep neural network (DNN). The remarkable representational power of DNN allows our framework to incorporate or approximate various extant data-driven models. We provide theoretical guarantees in terms of excess risk bounds for the DNN solution characterized by the network structure and sample size in a non-asymptotic manner, which justify the applicability of DNNs in the relevant contexts. Specifically, the convergence rate of the excess risk bound with respect to the sample size increases in the smoothness of the target quantile function but decreases in the dimension of feature variables. This rate can be further accelerated when the target function possesses a composite structure. In particular, our theoretical framework can be extended to accommodate the data-dependent scenarios, where the data-generating process could be time-dependent but not necessarily identical over time. Building on our theoretical results, we provide further managerial insights and practical guidance through simulation studies. Finally, we apply the DNN method to a real-world dataset obtained from a food supermarket. Our numerical experiments demonstrate that (1) the DNN method consistently outperforms other alternatives across a wide range of cost parameters, and (2) it exhibits good performance when the sample size is either very large or relatively limited.