DEBOSH: Deep Bayesian Shape Optimization
Keywords: gnns, optimization, shape, physics
TL;DR: Using Bayesian optimization to optimize 3D shapes using surrogate models
Abstract: Shape optimization is at the heart of many industrial applications, such as aerodynamics, heat transfer, and structural analysis. It has recently been shown that Graph Neural Networks (GNNs) can predict the performance of a shape quickly and accurately and be used to optimize shapes more effectively than traditional techniques. However, to fully explore the shape space, one must often explore shapes that deviate significantly from the training set. For these, GNN predictions become unreliable, something that is often ignored by most current methods. For classical optimization techniques, such as those relying on Gaussian Processes, Bayesian Optimization (BO) framework addresses this issue by enabling the model to assess its own prediction accuracy. Unfortunately, standard approaches to estimating neural network's uncertainty can entail long training times, high memory and computational requirements, and reduced model accuracy. In this paper, we propose a novel uncertainty-based method tailored to shape optimization that enables effective Bayesian optimization and increases the quality of the resulting shapes beyond that of popular state-of-the-art approaches.
Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 6521
Loading