2601.07834

#1 A Complete Decomposition of Stochastic Differential Equations [PDF²] [Copy] [Kimi¹] [REL]

Author: Samuel Duffield

We show that any stochastic differential equation with prescribed time-dependent marginal distributions admits a decomposition into three components: a unique scalar field governing marginal evolution, a symmetric positive-semidefinite diffusion matrix field and a skew-symmetric matrix field.

Subjects: Probability , Machine Learning , Statistics Theory

Publish: 2026-01-12 18:59:36 UTC

---