Sen Yang

Sen Yang

Ph.D., Operations Management, NYU Stern. Multi-LLM agent systems for quantitative research; online learning and stochastic optimization.

Publications & Working Papers

Working papers

Adaptive Gradient Descent Algorithms for Online Optimization Problems in Operations
Sen Yang, Jinzhi Bu, Siyi Wang. Working paper, 2024–. arXiv preprint forthcoming.

Develops adaptive online stochastic gradient descent algorithms achieving state-of-the-art sublinear regret in non-stationary environments — even when the magnitude of environmental variation is unknown a priori. Applications to multi-product inventory control, portfolio selection, and broader online convex optimization.

Online Gradient Descent Algorithm for Multi-Item E-commerce Order Fulfillment
Sen Yang, Divya Singhvi, Jiawei Zhang. Working paper, 2024–. arXiv preprint forthcoming.

A primal–dual online gradient descent algorithm for large-scale e-commerce fulfillment. The algorithm uses the gap between primal solutions and actual fulfillment decisions as the gradient signal for maintaining dual prices on inventory, sidestepping the need for a separate dual oracle.

Talks & teaching