OneCast: Structured Decomposition and Modular Generation for Cross-Domain Time Series Forecasting

TL;DR

OneCast introduces a structured, modular approach to cross-domain time series forecasting by explicitly decomposing series into seasonal and trend components, improving generalization across heterogeneous data.

Contribution

It proposes a novel framework that explicitly decouples and models seasonal and trend components using tailored generative pathways, enhancing cross-domain forecasting performance.

Findings

Outperforms state-of-the-art baselines on eight domains.

Effectively captures seasonal patterns with interpretable basis functions.

Successfully models domain-specific trends with a semantic-aware tokenizer.

Abstract

Cross-domain time series forecasting is a valuable task in various web applications. Despite its rapid advancement, achieving effective generalization across heterogeneous time series data remains a significant challenge. Existing methods have made progress by extending single-domain models, yet often fall short when facing domain-specific trend shifts and inconsistent periodic patterns. We argue that a key limitation lies in treating temporal series as undifferentiated sequence, without explicitly decoupling their inherent structural components. To address this, we propose OneCast, a structured and modular forecasting framework that decomposes time series into seasonal and trend components, each modeled through tailored generative pathways. Specifically, the seasonal component is captured by a lightweight projection module that reconstructs periodic patterns via interpretable basis functions. In parallel, the trend component is encoded into discrete tokens at segment level via a semantic-aware tokenizer, and subsequently inferred through a masked discrete diffusion mechanism. The outputs from both branches are combined to produce a final forecast that captures seasonal patterns while tracking domain-specific trends. Extensive experiments across eight domains demonstrate that OneCast mostly outperforms state-of-the-art baselines.