Sunday, December 7
Time – TBD
Additional Fee Required: $20
All participants are required to register for the 2025 Winter Simulation Conference. Sign up when you register for the meeting.
Light refreshments will be provided.
Overview
The SimOpt workshop introduces SimOpt – an open-source Python library of simulation models and optimization algorithms – and benchmarking platform. The workshop provides guidance on how to interact with the library. You will learn how to run multiple solvers on multiple problems and generate a range of diagnostic plots, such as the one below, that shed light on the relative performance of solvers, all with minimal effort. You will also learn how to build your own problems and solvers and use them within the platform to run simulation optimization experiments. The workshop also covers new SimOpt capabilities, such as data farming.
Who Should Attend?
The target audience for the workshop is researchers and advanced practitioners who are comfortable working with Python code.
Why Should You Attend?
You may have a simulation optimization problem and want to compare a few solvers to determine which is the best one to employ. You may be a solver developer and want to use SimOpt’s tools to compare your solver with others, aiming to improve its design. You may be a researcher of simulation methods in derivative estimation, metamodeling, random variate generation, variance reduction, etc., interested in applications for simulation optimization. You may be an educator looking for pedagogical tools or just be curious about simulation optimization.
SimOpt is hosted and sponsored by North Carolina State University, Cornell University and Texas A&M University.
The workshop is offered in-person only. Participation is limited and it is required that participants complete the installation instructions before attending the workshop.
Questions on the workshop can be directed to Sara Shashaani.