Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Feature to Dynamics: Feature-space to Autoregression strategy for Zero-shot Time Series Forecasting

About

Zero-shot time series forecasting aims to predict future values for previously unseen series, requiring models to generalize temporal dynamics beyond the training distribution. While recent foundation models achieve strong in-domain performance through large-scale pretraining, their effectiveness often relies on broad data coverage and implicit pattern memorization, which can limit generalization when data are scarce or source and target domains are disjoint. In this work, we propose FSA, a feature-to-strategy framework for controlled zero-shot univariate forecasting. Instead of directly modeling raw sequences in the observation space, FSA learns a structured mapping from an interpretable feature space to an autoregressive strategy space. This design introduces explicit inductive biases that disentangle global trends, periodic components, and local temporal dynamics, enabling the model to capture transferable time-series structure with fewer data assumptions. Empirical results show that, under identical pretraining data, training protocol, and comparable parameter budgets, FSA outperforms Transformer-based architectures in our controlled zero-shot setting.

Yifan Wu, Junjie Wu, Kai Wu, Xiaoyu Zhang, Jian Lou• 2026

Related benchmarks

TaskDatasetResultRank
Multivariate ForecastingETTh1
MSE0.456
830
Multivariate Time-series ForecastingETTm1
MSE0.562
686
Multivariate Time-series ForecastingETTm2
MSE0.121
539
Multivariate Time-series ForecastingWeather
MSE0.12
409
Multivariate Time-series ForecastingExchange
MAE0.126
262
Multivariate Time-series ForecastingETTh2
MSE0.198
198
Multivariate Time-series ForecastingElectricity
MAE0.416
105
Showing 7 of 7 rows

Other info

Follow for update