SurF: A Generative Model for Multivariate Irregular Time Series Forecasting
For practitioners working with asynchronous event streams, SurF provides a scalable generative approach that outperforms existing methods on multiple benchmarks, representing an initial step toward foundation models for this modality.
SurF introduces a generative model for multivariate irregular time series that leverages the Time Rescaling Theorem as a learnable bijection, enabling training across heterogeneous datasets. It achieves best reported time RMSE on three of six benchmarks and beats all baselines on two under a leave-one-out protocol.
Irregularly sampled multivariate event streams remain a stubbornly difficult modality for generative modeling: tokenization-based approaches break down when inter-event intervals vary by orders of magnitude, and neural temporal point processes are bottlenecked by window-level numerical quadrature. We (i) propose SurF, a generative model that uses the Time Rescaling Theorem (TRT) as a learnable bijection between event sequences and i.i.d.\ unit-rate exponential noise, enabling a single model to be trained across heterogeneous event-stream datasets; (ii) three efficient parameterizations of the cumulative intensity that scale to long sequences; and (iii) a Transformer-based encoder for multi-dataset pretraining. On six real-world benchmarks, SurF achieves the best reported time RMSE on Earthquake, Retweet, and Taobao, and is within trial-level noise of the strongest specialist on the remaining three. Under a strict leave-one-out protocol, the held-out checkpoint beats every classical and neural-autoregressive baseline on 5/6 datasets and beats every baseline on Amazon and Earthquake, an initial step toward foundation models over asynchronous event streams.