Advancing lattice data techniques using multiscale decomposition : simulations and applications

The supporting material includes figures and tables of the work where spatial lattice techniques and methods were developed to detect and predict spatial hotspots. The multiscale decomposition algorithm, the Discrete pulse transform, is first extended to irregular lattice data to detect hotspots in spatial lattice data. Thereafter the variogram estimation (adopted from geostatistical analysis) is extended to spatial lattice data which is then utilized in a Geographically and temporally weighted regression spatio-temporal kriging method to predict spatial hotspots.

本研究的辅助材料包含相关工作的图表与表格，该工作开发了用于检测与预测空间热点的空间格网技术与方法。首先将多尺度分解算法——离散脉冲变换（Discrete pulse transform）——拓展至不规则格网数据，以实现空间格网数据中的热点检测。随后将源自地统计分析的变异函数估计方法拓展至空间格网数据，并将其应用于地理时空加权回归时空克里金方法中，以完成空间热点的预测。