papersSEP 10 04:00 UTC
Paper introduces constraint-aware discrete black-box optimization using tensor decomposition
A new arXiv preprint tackles optimization of costly-to-evaluate functions over discrete domains under constraints. The proposed method incorporates tensor decomposition into sequential model-based optimization, using surrogate models to improve sample efficiency.