Xun Zhou

TL;DR: A Hetero-ConvL STM framework is proposed, where a few novel ideas are implemented on top of the basic ConvLSTM model, such as incorporating spatial graph features and spatial model ensemble, which makes reasonably accurate predictions and significantly improves the prediction accuracy over baseline approaches.

TL;DR: This survey reviews recent computational techniques and tools in spatiotemporal data mining and provides comprehensive coverage of computational approaches for various pattern families, focusing on several major pattern families.

TL;DR: This work considers a much more complexed scenario, in which multiple moving MDs sharing multiple heterogeneous MEC servers, and a problem named as minimum energy consumption problem in deadline-aware MEC system is formulated, and such problem is proved to be NP-hard.

TL;DR: In this paper, a multi-task learning (MTL) based approach was proposed to predict the hourly air pollution concentration on the basis of meteorological data of previous days by formulating the prediction over 24 hours as a multistask learning problem, and a useful regularization by enforcing the prediction models of consecutive hours to be close to each other.

TL;DR: The Capacity Constrained Route Planner (CCRP) is summarized, which generalizes shortest path algorithms by honoring capacity constraints and the spread of people over space and time and proposes a novel scalable algorithm that exploits the spatial structure of transportation networks to accelerate routing algorithms for large network datasets.