Temporal and spatial change detection for scientific data set stream

Identifying “important” time frames and sub-regions from massive scientific data set stream and directing scientists to pay attention to segments of interest in the data is an important research area. We introduced a general approach for change detection which based on statistical technology and designed algorithms to reveal important time-frames and sub-regions from scientific data set stream. Experiment Results obtained with synthetic data and plasma simulation data are presented. Our work can remarkably improve the way in which scientists extract useful information from large, complex, scientific data.