Physics-Informed Support Vector Kernels via Green-Function Analogies and Jackson-Chebyshev Spectral Design

The paper proposes a physics‑informed kernel design for support vector regression, using Green’s function analogies and Jackson‑Chebyshev spectral methods to construct a positive‑semidefinite kernel without requiring exact correspondence to a physical propagator. The resulting Jackson‑damped Chebyshev kernel provides an explicit feature map and a spectral prior tailored to structured observables. The authors benchmark the kernel on several physical regression tasks—including copper conductivity, Dirac‑like band dispersion, quartic‑oscillator energy levels, photonic‑crystal transmission, and Fibonacci‑chain transmission—using nested validation, learning curves, and comparisons to random‑forest, multilayer‑perceptron, and Nyström baselines.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.