STRATA-TS: Selective Knowledge Transfer for Urban Time Series Forecasting with Retrieval-Guided Reasoning
This addresses the challenge of reliable forecasting for cities with scarce data, offering an interpretable solution for urban planning applications, though it is incremental as it builds on existing transfer and retrieval methods.
The paper tackles the problem of data imbalance in urban time series forecasting by proposing STRATA-TS, a framework that selectively transfers knowledge from data-rich to data-scarce cities using retrieval-guided reasoning, achieving consistent performance improvements over baselines on parking availability datasets in Singapore, Nottingham, and Glasgow.
Urban forecasting models often face a severe data imbalance problem: only a few cities have dense, long-span records, while many others expose short or incomplete histories. Direct transfer from data-rich to data-scarce cities is unreliable because only a limited subset of source patterns truly benefits the target domain, whereas indiscriminate transfer risks introducing noise and negative transfer. We present STRATA-TS (Selective TRAnsfer via TArget-aware retrieval for Time Series), a framework that combines domain-adapted retrieval with reasoning-capable large models to improve forecasting in scarce data regimes. STRATA-TS employs a patch-based temporal encoder to identify source subsequences that are semantically and dynamically aligned with the target query. These retrieved exemplars are then injected into a retrieval-guided reasoning stage, where an LLM performs structured inference over target inputs and retrieved support. To enable efficient deployment, we distill the reasoning process into a compact open model via supervised fine-tuning. Extensive experiments on three parking availability datasets across Singapore, Nottingham, and Glasgow demonstrate that STRATA-TS consistently outperforms strong forecasting and transfer baselines, while providing interpretable knowledge transfer pathways.