Best Practices for Experiment Design with Marat Latypov, Paul Leu, and Vishwanath Hegadekatte

Опубликовано: 01 Июнь 2025
на канале: SigOpt

620

Models are only as effective as the design of experiments that evaluate, compare, test and validate them. Between real-world constraints, limitations of your computing environment, and everything in between, there are many decisions that need to be made to design an experiment to ask the right questions. These decisions and how they translate to more insightful experiments is the subject of this panel discussion.

Michael McCourt, Head of Engineering & Research at SigOpt, moderates a discussion that cuts across a variety of enterprise and university research applications to distill a set of foundational principles that can be universally applied to experiment design. Speakers include:

🔵 Marat Latypov, Assistant Professor of Materials Science and Engineering, University of Arizona
🔵 Paul Leu, Associate Professor, University of Pittsburgh
🔵 Vishwanath Hegadekatte, Global R&D Manager, Novelis

⏱️ 0:05 - Speaker intros
⏱️ 2:43 - How does SigOpt play a role in materials design
⏱️ 12:17 - How materials science has evolved from industry to academics
⏱️ 17:07 - What would help with the adoption of sample efficient exploration?
⏱️ 24:46 - Getting students access to tools like SigOpt
⏱️ 29:45 - How do you decide what the metrics of interest are?
⏱️ 39:52 - How do you work with limited observation budgets?

This panel was presented as a part of the 2021 SigOpt Summit.

✔️ Learn more about SigOpt: https://sigopt.com
✔️ Learn more about SigOpt Summit: https://summit.sigopt.com