Events

Prof. Zhuoyu (Daniel) Long
Professor | Department of Systems Engineering and Engineering Management
Department of Decisions, Operations and Technology
The Chinese University of Hong Kong

We study budget-constrained treatment allocation when heterogeneous treatment benefits and costs are estimated from data. We maximize worst-case welfare subject to a worst-case budget constraint over a Wasserstein ball centered at the plug-in empirical distribution of covariates and estimated scores. Our first contribution is a class-calibrated statistical analysis. Because score representations are not unique, we calibrate the nominal distribution to the class of score distributions with the correct covariate marginal and conditional mean benefit and cost, rather than to a fixed oracle score distribution. If the ambiguity set intersects this policy-valid class, the robust evaluations yield simultaneous causal welfare and budget certificates. For cross-fitted augmented inverse propensity weighted scores, the Wasserstein distance to this class is bounded by an empirical-measure term plus a doubly robust product of propensity-score and outcome-or cost-regression errors. Our second contribution solves the resulting infinite-dimensional functional optimization problem. Standard Wasserstein duality theory eliminates the distributional adversary but leaves optimization over all Borel-measurable policies infinite-dimensional. We show an exact sample-wise reduction: for randomized policies, a finite-dimensional linear program followed by a Shapley extension yields a globally optimal and robustly feasible full-space policy. For deterministic policies, a mixed-integer program and nearest-neighbor extension yield explicit welfare-loss and budget-excess bounds. Simulations and a JTPA case study show favorable out-of-sample welfare and budget control.

Daniel Zhuoyu Long is a Professor in the Department of Systems Engineering and Engineering Management at The Chinese University of Hong Kong. He received his Bachelor’s degree from Tsinghua University in 2005 Master’s degree from the Chinese Academy of Sciences in 2008 and Ph.D. from the National University of Singapore Business School in 2013 and joined CUHK in the same year. His research primarily focuses on distributionally robust optimization theory and its applications to various operations management problems including logistics and supply chain management project management healthcare operations management and revenue management. His work was selected as a finalist for the 2021 Best OM Paper Award in OR and the 2026 POMS CHOM Best Paper Competition. He currently serves as an AE for MSOM and on the editorial board of Engineering.