SAAPS contain several models for the prediction of the risk of an anomaly with a daily resolution. With this tool the model that best matches user submitted anomaly data can be found.
The best model is determined using linear correlation or conditional probability.
To find the best model the following steps must be performed:
Each model is run for the data entered as events and non-events and the statistics are computed. The result is displayed as an ordered list with the model giving the best agreement with the data at the top of the list. Selecting one model will display the result and information about the model.