MicroStrategy ONE

Predictive analysis overview

MicroStrategy provides a basic set of predictive analysis features. This analysis involves exploring an input dataset to discern relationships or patterns within the data. The results of this exploration, also referred to as training, can be used to create a predictive model which can then be applied to new datasets to provide predictions or forecasting.

In MicroStrategy, the following steps make up the predictive analysis process:

  1. Create a training metric.

  2. Add the training metric to a report.

  3. Execute the report. If specified in the Training Metric Wizard, this step results in the automatic creation of the predictive metric.

  4. If the predictive metric is not created automatically, create it manually.

The following is a list of the predictive analysis methods that MicroStrategy provides:

It is recommended that reports containing training metrics have report caching disabled. This is to ensure that the PMML model is always generated. Using report caching with reports that contain training metrics has the following effects:

  • If the report uses a report cache, there is no need for MicroStrategy to execute the training metric, since the results have already been cached. Therefore, the PMML model will not be generated.

  • If the training metrics are defined to automatically create predictive metrics, then the predictive metrics will be updated upon report execution.

While the predictive metrics are updated, due to the PMML model not being regenerated, the report results do not reflect any changes to the underlying data. To avoid this scenario by disabling report caching for a training report, double-click on the report to open it in the Report Editor. Then from the Data menu, select Report Caching Options. In the Report Caching Options dialog box that opens, disable report caching for the report.