Applied Research in Discrete Optimization
NEW COURSE · IE-GY 9113 A
Launching Fall 2026
Learn more during these upcoming Webinars:
Overview
Mixed-integer optimization powers some of the highest-stakes decisions in industry — scheduling hospitals, routing fleets, designing power grids, structuring portfolios. This graduate course offers a comprehensive study covering the theory of strong formulations; the core and large-scale algorithms, including cutting planes, branching, and machine learning techniques; and decomposition methods such as Benders, Lagrangian, and column generation. Through hands-on practice across healthcare, transportation, energy, and finance applications, students learn to turn hard combinatorial problems into deployable decision-support tools.
Curriculum
The curriculum integrates the full technological evolution — from ERP to RPA, Intelligent Automation, and autonomous Agentic AI — with a deep exploration of the managerial and human challenges that accompany it. Technology is not a standalone solution; it must be carefully orchestrated with organizational structures, human behavior, and leadership to drive real transformation. Through realworld case studies of success and failure, students gain a practical, evidence-based methodology for diagnosing, leading, and sustaining technology-driven change — structured around three core pillars. This learning is anchored by a hands-on project running across all 4 modules — where student teams build a real MVP Agentic System, layer by layer, and compete in a final showcase judged by industry partners.
Three Core Pillars
Formulations
Not every model of the same problem is equally good — some solve in seconds while mathematically equivalent ones run for hours. This module covers the core principles to develop strong formulations and understand their geometric structure.
Algorithms
Not every approach works for every problem. Solving discrete programs at scale demands a toolkit of powerful ideas such as cutting planes, branching, machine learning, Benders decomposition, lagrangian relaxation, and column generation.
Practice
Hands-on work using state-of-the-art industrial solvers and generative AI tools. Learning through real-world cases of large-scale implementations. Focus on evaluating and selecting solutions in practice.
Hands-on Project
Student teams choose a problem-driven track — solving a real-world decision problem — or a computational track, building and benchmarking an algorithm. Each team delivers a final presentation and report.
Who is this Course For?
Industrial Engineering
Students seeking rigorous mathematical foundations for tackling discrete decision problems in production, logistics, and systems design.
Management of Technology
Those students interested in operations or analytics who want the quantitative depth to lead data-driven decision-making at scale in general complex systems.
Financial Engineering, Mechanical Engineering, Civil and Urban Engineering
Future practitioners in operations-intensive industries — energy, infrastructure, transportation, finance — who need to formulate and solve large-scale constrained optimization problems.
Business School (NYU Stern)
The course complements Stern's core operations offerings and goes more in depth on the optimization foundations.
Faculty
