New framework discovers governing equations from multi-source datasets

Researchers have developed a new framework called MCO-PDE that can discover governing partial differential equations (PDEs) from multiple, distinct datasets. This competitive optimization approach trains separate neural networks for each dataset and then uses a weighting mechanism to assess dataset credibility and find a consensus. Integrated with a genetic algorithm for structural search, MCO-PDE can identify both the form and parameters of physical laws. The method has demonstrated success in recovering known equations with high accuracy using limited data and can handle complex, real-world scenarios, including wave-tank experiments. AI

IMPACT Enables more robust and accurate scientific discovery by leveraging heterogeneous data sources.

RANK_REASON The cluster contains a single academic paper detailing a new methodology. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• MCO-PDE

• paper
• other

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