L0MO
A functional Bayesian optimization method that conducts search over sparse subsets of a Reproducing Kernel Hilbert Space (RKHS). It achieves effective optimization in function space by simultaneously optimizing kernel locations and coefficients.
Timeline (last 90 days)
An arXiv paper proposes a new functional Bayesian optimization method called L0 Manifold Optimization (L0MO).
L0MO searches over a subset of functions with sparse representations in a Reproducing Kernel Hilbert Space (RKHS).
L0MO simultaneously optimizes two sets of quantities: the kernel locations and the kernel coefficients.
The L0 in the method name implies control over the number of nonzero terms in the representation.
The paper provides a unified analytical perspective discussing the relationship between L0MO and existing FBO methods.
The authors constructed a new set of FBO benchmark functions by porting standard finite-dimensional test functions to infinite-dimensional domains.
Experimental results demonstrate that L0MO comprehensively achieves superior performance over current state-of-the-art methods across a wide range of benchmarks.
As a preprint, L0MO's robustness on real-world large-scale problems, sensitivity to sparsity selection, and computational overhead of kernel function location optimization remain to be further examined.
All Facts (8)
As a preprint, L0MO's robustness on real-world large-scale problems, sensitivity to sparsity selection, and computational overhead of kernel function location optimization remain to be further examined.
50%UnverifiedL0MO searches over a subset of functions with sparse representations in a Reproducing Kernel Hilbert Space (RKHS).
50%UnverifiedL0MO simultaneously optimizes two sets of quantities: the kernel locations and the kernel coefficients.
50%UnverifiedThe L0 in the method name implies control over the number of nonzero terms in the representation.
50%UnverifiedThe paper provides a unified analytical perspective discussing the relationship between L0MO and existing FBO methods.
50%UnverifiedThe authors constructed a new set of FBO benchmark functions by porting standard finite-dimensional test functions to infinite-dimensional domains.
50%UnverifiedExperimental results demonstrate that L0MO comprehensively achieves superior performance over current state-of-the-art methods across a wide range of benchmarks.
50%UnverifiedAn arXiv paper proposes a new functional Bayesian optimization method called L0 Manifold Optimization (L0MO).
50%