A new research paper proposes a method for generating sparse designs in high-dimensional computer experiments by quantizing the uniform measure on hypercubes. This approach addresses the limitations of traditional maximin and minimax criteria, which struggle with high dimensions and boundary sensitivity. The proposed method leverages Zador's theorem for a smooth transition to uniform designs and demonstrates that designs supported on fractional factorial subsets of hypercube vertices perform exceptionally well. AI
RANK_REASON The cluster contains a single academic paper published on arXiv. [lever_c_demoted from research: ic=1 ai=0.4]
- arXiv
- arXiv:2610.11772
- cs.LG
- Hugging Face
- Sparse designs via quantization of the uniform measure on high-dimensional hypercubes
- stat.ML
- Zador's celebrated theorem
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →