Search papers, labs, and topics across Lattice.
This paper introduces TimeBlocks, a novel framework for constructing lightweight, modular time-series models that can efficiently process real-time data streams. By leveraging a pool of interchangeable model blocks and a routing strategy, TimeBlocks adapts to varying conditions and tasks while maintaining continual calibration through a method called StreamCore. Experimental results demonstrate that TimeBlocks significantly outperforms existing foundational time-series models, addressing the limitations of size and adaptability in dynamic environments.
TimeBlocks outperforms traditional time-series models by dynamically constructing lightweight, adaptable models that excel in real-time processing.
The ongoing digitization has led to a proliferation of time-series data streams that monitor a variety of processes, from which valuable insights may be obtained. Further, the emergence of successful foundational language models begs the question of whether it is possible to achieve time-series models with the foundational properties of handling multiple tasks, while being sufficiently lightweight to allow real-time data stream processing. Existing foundational time-series models are often large and only effective in offline settings without stringent time and computational constraints, and where repeated model calibration is not needed. However, when applied to data streams, these models are ineffective due to their size and lack of support for continual calibration, which compromise their ability to deliver accurate real-time responses, their durability, and their deployability in hardware-limited settings. We propose TimeBlocks to enable versatile time-series processing by facilitating the efficient building of lightweight models suitable for multiple tasks under variable conditions. In particular, the method maintains a pool of interchangeable and modular model blocks that can be used to construct new time-series models. When presented with specific time-series data, a routing strategy iteratively selects the most suitable blocks to construct a lightweight and accurate model for the data. We equip TimeBlocks with a method called StreamCore to build a representative small subset of the data stream, which preserves a guaranteed approximation of the stream over time, enabling continual model calibration. An experimental study on multiple data sets and covering multiple tasks shows that TimeBlocks enables to build models capable of outperforming existing baselines.
