New Scientific ML Method Learns Chaotic System Equations

Researchers have developed PEM-UDE, a novel scientific machine learning method designed to extract interpretable mathematical models from complex, chaotic dynamical systems, even when faced with noisy experimental data. This technique combines prediction-error methodology with universal differential equations to smooth the optimization process and identify governing equations. The method was successfully tested on benchmark chaotic systems like the Rössler attractor and a real electrical circuit, accurately recovering functional forms from noisy observations. Furthermore, PEM-UDE was applied to learn a multi-scale neural mass model from a population of Izhikevich neurons, linking single-neuron parameters to macroscopic network dynamics and predicting trends in oscillation frequency and synchrony. AI

IMPACT This method could enable more accurate and interpretable modeling of complex biological and physical systems.

RANK_REASON The cluster contains a scientific paper detailing a new machine learning method for modeling chaotic systems. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Helmut H Strey
• Hugging Face
• Izhikevich neurons
• PEM-UDE
• Rössler attractor

• paper
• model release

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