Universal Time Series Generation with Neural Controlled Differential Equations
Authors: Torben Berndt, Elyes Farjallah, Leif Seute, Raeid Saqur, Benjamin Walker, Jan Stühmer
Organizations: Heidelberg Institute for Theoretical Studies, Heidelberg, Germany · IAR, Karlsruhe Institute of Technology, Karlsruhe, Germany · Max Planck Institute for Polymer Research, Mainz, Germany · IWR, Heidelberg University, Heidelberg, Germany · Dept. of Computer Science, University of Toronto, Toronto, Canada · Mathematical Institute, University of Oxford, Oxford, UK · Vector Institute, Toronto, Canada
Recent work on the sequence universality of State Space Models (SSMs) has introduced efficient, maximally expressive continuous-time approaches for time-series modelling. While these works focus on discriminative settings, we extend this perspective to generative time-series modelling by proving that maximally expressive Structured Linear Controlled Differential Equations (SLiCEs) are universal time-series generators, in the sense that they can approximate the induced path laws of continuous causal pushforwards on compact latent sets in W∞. Building on these theoretical results, we propose Generative SLiCEs (G-SLiCEs), a maximally expressive continuous-time model for flow matching on path-space. Empirically, we show that expressivity improves performance in probabilistic forecasting and downstream tasks, while retaining the advantages of continuous-time models such as generalising to arbitrary observation grids. This is particularly beneficial for irregular grids, where fixed-grid models often struggle.