BO-pro-c-tr reduces regret and overhead in large-scale Bayesian optimization
Achieves 2.6% reduction in simple regret and 34.9% in computational overhead compared to independent local GP models.
Stories per day · 30 days
1 RSS
Relatedgaussian-process2acquisition-functions1black-box-optimization1cma-es1combinatorial-optimization1freeze-thaw1high-dimensional1hyperparameter-optimization1information-directed-sampling1sequential-decision-making1synthetic-biology1thompson-sampling1trust-regions1
10 stories, newest first · one card per story
Stories tagged #bayesian-optimization
EI Stopping Rule for Cost-Aware Freeze-Thaw Bayesian Optimization
One-step EI stopping balances expected improvement against evaluation costs in freeze-thaw Bayesian optimization.
New acquisition speeds Bayesian optimization on the unit cube
The proposed acquisition is globally solvable faster than multi-start quasi-Newton and matches Gaussian-process accuracy per evaluation on cheaper black boxes.
Paper amortizes Thompson sampling for simulator-based design
Precomputed approximations enable information-directed sampling and are shown on two synthetic-biology simulators with intractable likelihoods.
Fast Bayesian optimization methods match SOTA at 100x–1000x speed
The methods target continuous and combinatorial inputs from about 10 to more than 6000 dimensions by using linear structure already in BO.
AIM: Agentic Idea Management for Automated Research
AIM manages research ideas with Bayesian-style selection, beating AutoLab baselines by up to 4.9 points and converging 3.1x faster.
Context-Continuous Preference Learning for Exoskeleton Personalization
A Gaussian-process preference model shares nearby-context feedback to personalize ankle and elbow exoskeleton assistance with less data.
JAREX: An Acquisition Function for Multi-Objective Algorithmic Process Characterization
JAREX maps multi-objective pharmaceutical process pass regions with fewer experiments than factorial design.
Learning to Move Cities: Deep Meta-Models and Reinforcement Policies for Calibration and Control in Urban Networks
Shared latent traffic models calibrate urban simulators and cut travel time by up to 51% with deep Q-learning.
Reduced-Space Multi-Fidelity Bayesian Optimization of Process Simulation Models
RS-MFBO couples global sensitivity analysis with fidelity-augmented Gaussian processes to slash costly high-fidelity simulation runs in industrial flowsheet optimization.