Research map / Probabilistic and scientific modeling / Bayesian optimization
ALAS: Additive Learnable Alpha-Stable Kernels for Flexible Bayesian Optimization
Weibo Huang, Cheng Hua
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| learn the stability parameter alpha instead of RBF fixed Gaussian smoothness (ablation) | GP with 1D ALAS component · Weierstrass function (n=25, single run) | average test PLL: -0.16 -> 0.23 | Improved |
| learn alpha on a smooth function sampled from an RBF GP (ablation) | GP with 1D ALAS component · Smooth RBF-GP sample (n=25) | average test PLL: RBF 2.791 vs ALAS 2.777 | No clear effect |
| use a single d-dimensional ALAS kernel with shared alpha instead of the separable ALAS-Sep (ablation) | ALAS · Rosenbrock-10d, Robot-4d, Portfolio-5d | log optimality gap | Improved |
| use separable ALAS-Sep with per-dimension alpha_j (ablation) | ALAS · Weierstrass-3d, Levy-20d, Rastrigin-30d | log optimality gap | Improved |
| learn an additional alpha parameter during early BO iterations (ablation) | ALAS · several BO benchmarks | log optimality gap | Hurt |
| learn alpha on smoother benchmarks instead of a fixed Gaussian kernel | ALAS · Exponential-5d, Hartmann-3d/6d | log optimality gap: alpha converges to 1.96, 1.86, 1.92 | No clear effect |
Closest papers by meaning
- BARK: A Fully Bayesian Tree Kernel for Black-box Optimization (ICML 2025 · similarity 0.88)
- Automated Kernel Discovery Towards Understanding High-dimensional Bayesian Optimization (NeurIPS 2026 · similarity 0.87)
- Kernel-based guarantees for nonlinear parametric models in Bayesian optimization (NeurIPS 2026 · similarity 0.87)
- Adaptive Kernel Design for Bayesian Optimization Is a Piece of CAKE with LLMs (NeurIPS 2025 · similarity 0.87)
- Robust and Computation-Aware Gaussian Processes (NeurIPS 2025 · similarity 0.86)
- Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimization (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
