COBALT performs direct discrete optimization over high-dimensional categorical structural designs by anchoring latent embeddings as graphs and applying trust-region acquisition on additive Gaussian process surrogates fitted to Monte Carlo finite-element data.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Categorical Optimization with Bayesian Anchored Latent Trust Regions for Structural Design under High-Dimensional Uncertainty
COBALT performs direct discrete optimization over high-dimensional categorical structural designs by anchoring latent embeddings as graphs and applying trust-region acquisition on additive Gaussian process surrogates fitted to Monte Carlo finite-element data.