{"id":"arxiv/2605.24548","url":"https://arxiv.org/abs/2605.24548","tags":["cs.LG","math.PR"],"score":0,"title":"Deep ZakaiJ: Structured Filtering for Jump-Diffusion Time Series Forecasting","domain":"cs.LG","source":"arxiv","authors":["Yan Leng","Thibaut Mastrolia","Hao Wang"],"reviews":0,"subject":"cs.LG","summary":"Time series driven by unobserved latent states frequently exhibit abrupt jump discontinuities whose timing and magnitude cannot be predicted from observed history alone. Classical jump-diffusion models offer a principled mathematical framework but assume rigid parametric forms, while recent neural jump models operate on fully observed trajectories without inferring the hidden states that govern the dynamics. We propose \\textit{Deep ZakaiJ}, a latent-state model for partially observed jump-diffusion systems that embeds the Zakai nonlinear filtering equation into a neural encoder--decoder…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-05-23","reactions":{}}