Abstract:Modeling and inferring spatial relationships and predicting missing values of environmental data are some of the main tasks of geospatial statisticians. These routine tasks are accomplished using multivariate geospatial models and the cokriging technique. The latter requires the evaluation of the expensive Gaussian log-likelihood function, which has impeded the adoption of multivariate geospatial models for large multivariate spatial datasets. However, this large-scale cokriging challenge provides a fertile ground for supercomputing implementations for the geospatial statistics community as it is paramount to scale computational capability to match the growth in environmental data coming from the widespread use of different data collection technologies. In this paper, we develop and deploy large-scale multivariate spatial modeling and inference on parallel hardware architectures. To tackle the increasing complexity in matrix operations and the massive concurrency in parallel systems, we leverage low-rank matrix approximation techniques with task-based programming models and schedule the asynchronous computational tasks using a dynamic runtime system. The proposed framework provides both the dense and the approximated computations of the Gaussian log-likelihood function. It demonstrates accuracy robustness and performance scalability on a variety of computer systems. Using both synthetic and real datasets, the low-rank matrix approximation shows better performance compared to exact computation, while preserving the application requirements in both parameter estimation and prediction accuracy. We also propose a novel algorithm to assess the prediction accuracy after the online parameter estimation. The algorithm quantifies prediction performance and provides a benchmark for measuring the efficiency and accuracy of several approximation techniques in multivariate spatial modeling.

Design and implementation of a parallel geographically weighted -nearest neighbor classifier.

Design and implementation of a parallel geographically weighted k-nearest neighbor classifier

ParSymG: a Parallel Clustering Approach for Unsupervised Classification of Remotely Sensed Imagery

Adaptable Parallel Strategy To Extract Polygons From Massive Classified Images On Multi-Core Clusters

A spatial Gaussian process method for hyperspectral remote sensing imagery classification

A parallel implementation of nearest neighbor analysis based on GPGPU

Parallel Calculation and Efficiency Analysis for Neighborhood Statistic Algorithm in Digital Terrain Analysis

Parallel Geospatial Analysis on Windows HPC Platform

High Performance Multivariate Geospatial Statistics on Manycore Systems

A new method of parallel sar image classification

Parallelization of Spectral Clustering Algorithm on Multi-Core Processors and GPGPU

Decomposition Method of Raster Geographic Data Based on Parallel Computing

A Parallel Varied Density-Based Clustering Algorithm with Optimized Data Partition

PGNN-Net: Parallel Graph Neural Networks for Hyperspectral Image Classification Using Multiple Spatial-Spectral Features

Optimizing Spatial Relationships in GCN to Improve the Classification Accuracy of Remote Sensing Images

The geographical weighted K-NN classifiers in land cover classification from remote sensing image: A case study of a subregion of Xi'an, China

GPU-based Acceleration of the Hyperspectral Band Selection by SNR Estimation Using Wavelet Transform

A New Algorithm for Large-Scale Geographically Weighted Regression with K-Nearest Neighbors

Self-adapted Genetic Hyperplane Classifier Algorithm for Multi-dimensional Remote Sensing Image

Parallel Approaches to Neighborhood Rough Sets: Classification and Feature Selection

Accelerating Geospatial Analysis on GPUs Using CUDA