A novel spatio-temporal data analysis approach based on prospective support vector clustering

Spatio-temporal data mining has recently gained considerable attention from both the research and practitioner communities. In this paper, we propose a novel spatio-temporal data analysis approach which is aimed at discovering abnormal spatial clustering patterns in a timely manner. Our approach is based on a robust clustering engine using support vector machines and incorporates the ideas from existing online surveillance methods to monitor for incremental changes over time. Two simulated scenarios are created to evaluate our approach. Initial experimental results indicate that our approach is able to detect abnormal areas with irregular shapes faster and more accurately than a widely-used spatio-temporal data analysis approach based on scan statistics.