[2508.17521] Modeling Irregular Astronomical Time Series with Neural Stochastic Delay Differential Equations

Computer Science > Machine Learning arXiv:2508.17521 (cs) [Submitted on 24 Aug 2025 (v1), last revised 2 Apr 2026 (this version, v2)] Title:Modeling Irregular Astronomical Time Series with Neural Stochastic Delay Differential Equations Authors:YongKyung Oh, Seungsu Kam, Dong-Young Lim, Sungil Kim View a PDF of the paper titled Modeling Irregular Astronomical Time Series with Neural Stochastic Delay Differential Equations, by YongKyung Oh and 3 other authors View PDF HTML (experimental) Abstract:Astronomical time series from large-scale surveys like LSST are often irregularly sampled and incomplete, posing challenges for classification and anomaly detection. We introduce a new framework based on Neural Stochastic Delay Differential Equations (Neural SDDEs) that combines stochastic modeling with neural networks to capture delayed temporal dynamics and handle irregular observations. Our approach integrates a delay-aware neural architecture, a numerical solver for SDDEs, and mechanisms to robustly learn from noisy, sparse sequences. Experiments on irregularly sampled astronomical data demonstrate strong classification accuracy and effective detection of novel astrophysical events, even with partial labels. This work highlights Neural SDDEs as a principled and practical tool for time series analysis under observational constraints. Comments: Subjects: Machine Learning (cs.LG) Cite as: arXiv:2508.17521 [cs.LG]   (or arXiv:2508.17521v2 [cs.LG] for this version)   https://doi.org/...