Search papers, labs, and topics across Lattice.
The paper introduces LAGO, a novel optimization framework that adaptively combines gradient-enhanced Bayesian Optimization (BO) for global exploration with trust region methods for local refinement. LAGO decouples global and local search by optimizing the BO acquisition function outside the active trust region and selectively incorporating local gradient information into the global Gaussian process based on a lengthscale criterion. This approach balances global exploration and local exploitation, leading to improved performance compared to standard local optimization algorithms while maintaining fast local convergence.
LAGO offers a way to get the best of both worlds in optimization, combining the global search of Bayesian Optimization with the fast local convergence of trust region methods through an adaptive competition that avoids numerical instability.
We introduce LAGO, a LocAl-Global Optimization algorithm that combines gradient-enhanced Bayesian Optimization (BO) with gradient-based trust region local refinement through an adaptive competition mechanism. At each iteration, global and local optimization strategies independently propose candidate points, and the next evaluation is selected based on predicted improvement. LAGO separates global exploration from local refinement at the proposal level: the BO acquisition function is optimized outside the active trust region, while local function and gradient evaluations are incorporated into the global gradient-enhanced Gaussian process only when they satisfy a lengthscale-based minimum-distance criterion, reducing the risk of numerical instability during the local exploitation. This enables efficient local refinement when reaching promising regions, without sacrificing a global search of the design space. As a result, the method achieves an improved exploration of the full design space compared to standard non-linear local optimization algorithms for smooth functions, while maintaining fast local convergence in regions of interest.
