SparseST: Exploiting Data Sparsity in Spatiotemporal Modeling and Prediction

2025-11-18

AI Summary

Existing spatiotemporal forecasting models incur prohibitive computational overhead and suffer from severe data and feature redundancy on resource-constrained edge devices. Method: This paper proposes the first efficient spatiotemporal modeling framework explicitly leveraging data sparsity. It introduces a sparsely activated ConvLSTM architecture, a data-sparsity-aware training strategy, and a redundant-feature filtering mechanism; further, it employs a multi-objective composite loss function to jointly optimize accuracy and efficiency, enabling customizable Pareto-optimal trade-offs. Contribution/Results: It is the first work to explicitly incorporate data sparsity modeling into the spatiotemporal forecasting pipeline. Evaluated on real-world edge tasks—including traffic flow and medical time-series prediction—the framework achieves an average 2.3× inference speedup and 47% memory reduction over baseline models, while matching state-of-the-art accuracy. This establishes a novel paradigm for lightweight spatiotemporal intelligence.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Spatiotemporal data mining (STDM) has a wide range of applications in various complex physical systems (CPS), i.e., transportation, manufacturing, healthcare, etc. Among all the proposed methods, the Convolutional Long Short-Term Memory (ConvLSTM) has proved to be generalizable and extendable in different applications and has multiple variants achieving state-of-the-art performance in various STDM applications. However, ConvLSTM and its variants are computationally expensive, which makes them inapplicable in edge devices with limited computational resources. With the emerging need for edge computing in CPS, efficient AI is essential to reduce the computational cost while preserving the model performance. Common methods of efficient AI are developed to reduce redundancy in model capacity (i.e., model pruning, compression, etc.). However, spatiotemporal data mining naturally requires extensive model capacity, as the embedded dependencies in spatiotemporal data are complex and hard to capture, which limits the model redundancy. Instead, there is a fairly high level of data and feature redundancy that introduces an unnecessary computational burden, which has been largely overlooked in existing research. Therefore, we developed a novel framework SparseST, that pioneered in exploiting data sparsity to develop an efficient spatiotemporal model. In addition, we explore and approximate the Pareto front between model performance and computational efficiency by designing a multi-objective composite loss function, which provides a practical guide for practitioners to adjust the model according to computational resource constraints and the performance requirements of downstream tasks.

Problem

• Reducing computational cost of spatiotemporal models for edge devices
• Addressing data and feature redundancy in spatiotemporal data mining
• Balancing model performance with computational efficiency constraints

Innovation

• Exploiting data sparsity for efficient spatiotemporal modeling
• Designing multi-objective loss for performance-efficiency tradeoff
• Reducing computational burden through feature redundancy elimination