A high-dimensional BO algorithm maximizes acquisition functions over discrete low-dimensional subspaces without assuming low-dimensional structure on the objective, achieving sub-linear regret with a tunable bound that removes the sqrt(D) factor for large subspace counts.
H.; Bindel, D.; and Wilson, A
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Trading Convergence Rate with Computational Budget in High Dimensional Bayesian Optimization
A high-dimensional BO algorithm maximizes acquisition functions over discrete low-dimensional subspaces without assuming low-dimensional structure on the objective, achieving sub-linear regret with a tunable bound that removes the sqrt(D) factor for large subspace counts.