arXiv:2606.08479v1 Announce Type: new Abstract: Inferring the forces that drive a dynamical system from partial observations is a fundamental challenge across physics, particularly when distinct underlying mechanisms produce similar observable dynamics. Here we show that the effective muscular forcing underlying avian respiratory dynamics can be reconstructed from measurements of air-sac pressure alone. Using an interpretable learning framework based on Kolmogorov-Arnold networks, we infer the governing equations of the system directly from data and uncover a nontrivial structure in the underl
The increasing maturity of interpretable machine learning frameworks like Kolmogorov-Arnold networks allows for the extraction of governing equations from complex biological data, addressing a long-standing challenge in understanding dynamical systems.
This development offers a new methodology for reverse-engineering biological processes and other complex systems directly from observational data, which could accelerate discovery and control in fields like medicine and engineering.
The ability to infer hidden forcing mechanisms and governing equations from partial observations using interpretable AI changes how researchers can approach the analysis of complex dynamical systems, moving beyond purely correlational insights to causal inference.
- · Biomedical researchers
- · AI model developers (interpretable AI)
- · Pharmaceutical industry
- · Robotics and control systems engineers
- · Traditional empirical modeling approaches (some aspects)
- · Purely black-box AI applications
This method could lead to more robust and predictive models of biological systems.
It may enable novel therapeutic interventions or bio-inspired engineering designs by uncovering fundamental principles.
Widespread adoption could lead to a paradigm shift in scientific discovery, integrating AI for hypothesis generation and causal mechanism identification across various fields.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG