Event Details

April 23, 2026

Join Zoom Webinar

Passcode: 345051

Webinar ID: 964 8206 8919

Abstract

One of the key challenges in decision-focused learning is that for many problems of interest (like combinatorial optimization with an uncertain objective), the decision-loss is ill-behaved, piecewise constant, and NP-Hard to optimize. Consequently, numerous authors have proposed various surrogate loss functions to approximate the decision loss, often based on ad hoc approximations and heuristic reasoning. In this talk, we present a unifying and principled framework for these surrogate losses. Specifically, we show that the (ill-behaved) decision loss is equivalent to a particular directional gradient of a plug-in functional, and we can derive an entire family of surrogate losses by invoking different zeroth-order approximations for this gradient. We call the resulting surrogates “Perturbation Gradient” losses, and provide a difference-of-convex functions representation. Several popular losses are either special cases of a perturbation gradient loss, or a particular convexification of a perturbation gradient loss (e.g., SPO+, Fenchel-Young Loss, Differentiable Black Box Optimizers). Leveraging this relationship, we can provide novel performance guarantees for many of these methods, and, most importantly, suggest simple, computationally cheap modifications to the original losses that lead to substantive performance improvements empirically.

Host: Eric Boxer

POC: Justina Gilleland

Speaker Bio

Vishal Gupta is an Associate Professor of Data Sciences and Operations at the USC Marshall School of Business and the Dean's Associate Professor of Business Administration. Because of his research interests and expertise, he also holds a courtesy appointment in USC Viterbi’s School of Engineering in Industrial and Systems Engineering, and is an affiliate faculty with USC’s Center for AI and Society.

Before joining USC, Vishal Gupta completed his B.A. in Mathematics and Philosophy at Yale University, graduating Magna Cum Laude with honors, and completed Part III of the Mathematics Tripos at the University of Cambridge with distinction. He then spent four years working as a “quant” in finance at Barclays Capital, focusing on commodities modeling, derivatives pricing, and risk management.

Eventually, Vishal realized how much he missed working towards a larger mission of impact, and left the private sector to complete his Ph.D. in Operations Research at MIT in 2014.

Vishal’s research focuses on data-driven decision-making and optimization, particularly in settings where data are scarce.
Such settings are common in applications that rely on personalization (like precision healthcare) and real-time decision-making (like risk management). Consequently, his research spans a wide variety of areas including education, healthcare, and applications of artificial intelligence intelligence. Vishal has received a number of recognitions for his work, including the Wagner Prize for Excellence in the Practice of Advanced Analytics and Operations Research, the Pierskalla Best Paper Prize, the Jagdish Sheth Impact of Research on Practice Award.