New method identifies optimal configurations across multiple computer experiments

Researchers have developed a new method called "joint contour location" (jCL) to efficiently identify input configurations that yield specific outcomes across multiple computer experiments simultaneously. This approach is particularly useful for complex scenarios, such as finding stable flight conditions that result in zero torque forces in vehicle dynamics. The jCL scheme balances exploration of response surfaces with exploitation of learned intersecting contours, employing Gaussian processes (GPs), multitask GPs, and deep GPs. This novel method significantly outperforms existing contour location strategies and optimization-based alternatives. AI

IMPACT This new method could improve the efficiency of complex simulations and optimization tasks in various scientific and engineering fields.

RANK_REASON The cluster contains an academic paper detailing a new methodology for computer experiments. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.LG →

• Actively Learning Joint Contours of Multiple Computer Experiments
• Annie S. Booth
• arXiv
• Deep GPs
• Gaussian Processes
• multitask GPs

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →