A Predictive Analytics Framework for Spatiotemporal Hotspots

Project Start Date: Jul 1, 2014
Research Areas: Analytics, Analytics - Machine Learning, Analytics - Probabilistic Modeling
Funding: Member Funded
Project Tags: , ,

Project Summary

Data mining’s future is predictive analytics. We will develop a predictive analytics framework for spatio-temporal data utilizing state-of-the-art big data technologies. In spatiotemporal data, people are interested in hotspots, areas of space and time with unusually high incidences of events. You may also want to predict how these hotspots may grow and where future hotspots may occur. To facilitate such analyzing and forecasting needs, this project has two specific objectives:

1) developing hotspots detection tool applying spatiotemporal clustering techniques

2) empowering hotspots forecasting through evolutionary clustering, time series prediction and geospatial prediction Hotspots analysis and prediction is very useful in public health, homeland security, auditing selection, and business & marketing.