With the goal of predicting the future rainfall intensity in a local region over a relatively short period time, precipitation nowcasting has been a long-time scientific challenge with great social and economic impact. The radar echo extrapolation approaches for precipitation nowcasting take radar echo images as input, aiming to generate future radar echo images by learning from the historical images. To effectively handle complex and high non-stationary evolution of radar echoes, we propose to decompose the movement into optical flow field motion and morphologic deformation. Following this idea, we introduce Flow-Deformation Network (FDNet), a neural network that models flow and deformation in two parallel cross pathways. The flow encoder captures the optical flow field motion between consecutive images and the deformation encoder distinguishes the change of shape from the translational motion of radar echoes. We evaluate the proposed network architecture on two real-world radar echo datasets. Our model achieves state-of-the-art prediction results compared with recent approaches. To the best of our knowledge, this is the first network architecture with flow and deformation separation to model the evolution of radar echoes for precipitation nowcasting. We believe that the general idea of this work could not only inspire much more effective approaches but also be applied to other similar spatio-temporal prediction tasks.

Urban sensing is one of the fundamental building blocks of urban computing. It uses various types of sensors deployed in different geospatial locations to continuously and cooperatively monitor the natural and cultural environment in urban areas. Nevertheless, issues such as uneven distribution, low sampling rate and high failure ratio of sensors often make their readings less reliable. This paper provides an innovative framework to detect the noise data as well as to repair them from a spatial-temporal causality perspective rather than to deal with them individually. This can be achieved by connecting data through monitored objects, using the Skip-gram model to estimate spatial correlation and long short-term memory to estimate temporal correlation. The framework consists of three major modules: 1) a space embedded Bidirectional Long Short-Term Memory (BiLSTM)-based sequence labeling module to detect the noise data and the latent missing data; 2) a space embedded BiLSTM-based sequence predicting module calculating the value of the missing data; 3) an object characteristics fusion repairing module to correct the spatial and temporal dislocation sensory data. The approach is evaluated with real-world data collected by over 3000 electronic traffic bayonet devices in a citywide scale of a medium-sized city in China, and the result is superior to those of several referenced approaches. With a 12.9% improvement in data accuracy over the raw data, the proposed framework plays a significant role in various real-world use cases in urban governance, such as criminal investigation, traffic violation monitoring, and equipment maintenance.

With the advent of the big data era, the amounts of sampling data and the dimensions of data features are rapidly growing. It is highly desired to enable fast and efficient clustering of unlabeled samples based on feature similarities. As a fundamental primitive for data clustering, the k-means operation is receiving increasingly more attentions today. To achieve high performance k-means computations on modern multi-core/many-core systems, we propose a matrix-based fused framework that can achieve high performance by conducting computations on a distance matrix and at the same time can improve the memory reuse through the fusion of the distance-matrix computation and the nearest centroids reduction. We implement and optimize the parallel k-means algorithm on the SW26010 many-core processor, which is the major horsepower of Sunway TaihuLight. In particular, we design a task mapping strategy for load-balanced task distribution, a data sharing scheme to reduce the memory footprint and a register blocking strategy to increase the data locality. Optimization techniques such as instruction reordering and double buffering are further applied to improve the sustained performance. Discussions on block-size tuning and performance modeling are also presented. We show by experiments on both randomly generated and real-world datasets that our parallel implementation of k-means on SW26010 can sustain a double-precision performance of over 348.1 Gflops, which is 46.9% of the peak performance and 84% of the theoretical performance upper bound on a single core group, and can achieve a nearly ideal scalability to the whole SW26010 processor of four core groups. Performance comparisons with the previous state-of-the-art on both CPU and GPU are also provided to show the superiority of our optimized k-means kernel.