Spatio-temporal predictive analytics, which aims to analyze and model data with both spatial and temporal information for future prediction and forecasting, has long attracted extensive research interest and has assumed increasing importance as the massive increases in both data availability and data diversity make spatio-temporal data more ubiquitous than ever. The problem of modeling spatio-temporal data is extremely important but yet quite challenging, primarily due to (1) substantial data heterogeneity in terms of data formats, sizes, and resolutions; (2) complex data dependency at varying spatiotemporal scales within and/or across heterogeneous data sources; and (3) various types of noise in the heterogeneous spatio-temporal data.
This project represents a bold attempt at tackling the aforementioned fundamental challenges. We will develop a novel STL framework that effectively integrates, models, and analyzes the spatio-temporal data for predictive analytics. Under the STL framework, we will further develop a number of new methods for effectively integrating data from heterogeneous sources while preserving the intrinsic spatio-temporal structure of data from each source via tensor-based operations, for quantitatively characterizing, effectively learning, and faithfully interpreting the complex dependency of heterogeneous spatio-temporal data, as well as for addressing various sources of data noise for robust spatio-temporal predictive analytics, respectively.
The developed framework and methods will be general in nature, contributing to both machine learning foundations and data analytics techniques. At the same time, the results obtained from this project will also have direct impacts on solving real-world challenging problems in spatio-temporal predictive analytics, as encountered in epidemiology, transportation, and climate science.
This project is supported by the Research Grants Council (RGC), Hong Kong SAR, China (Project HKBU 211212, 12202415, and 12201318).
For further information on this research topic, please contact Prof. Jiming, LIU.