VirusStrategy Predictive Analytics uses mathematical algorithms to assimilate geospatial time series data and make forward predictions.

• Generalised model – not domain specific. It can potentially apply to any spatial event data (e.g. crime, earthquakes, consumer behaviour, risk, epidemiology etc.).
• State of the art approach - performs nonparametric estimation of intensities at the same time as colouring the events.
• Clever optimisation - traditional models will not scale to large datasets because computational complexity is a function of the square of the number of events. VirusStrategy uses clever optimisation to reduce the processing time and uses simple geometry as an approximation.
• Enables the use of other numeric event characteristics to be included in the  model. These covariates could be anything upon which Euclidean distance is  meaningful e.g. the magnitude of the tremor, the monetary value of the purchase, the number of police on patrol etc. All of these things may have a bearing on the  two intensity functions.
• An arbitrary number of covariates may be used at linear computational cost.

2009 H1N1: Actual last 30 days

2009 H1N1: Predicted last 30 days

"Contact us to download the software today."