Prerequisites

• Previously created predictive analysis type.
• SoftExpert Performance among the solutions acquired by your organization.
• Access to the Predictive analysis (DL001) menu.

Introduction

With a predictive analysis type created, it is possible to add a KPI forecasting analysis.

KPI forecasting applies time series techniques, machine learning, and statistics to anticipate future values ​​of performance indicators, on the basis of historical data.

The analytical engine executes 28 algorithms, calculates error metrics, and automatically presents the 10 most accurate models, highlighting the winner with its projections.

Data can be obtained directly from SoftExpert Performance or from Excel spreadsheets, ensuring flexibility in the origin of the time series.

Creating an analysis

See below how to add a KPI forecasting analysis:

1. Access the Predictive analysis (DL001) menu.

2. Expand the Create button and select the KPI Forecasting option.

3. On the screen that will open, enter an ID # and a name for the record, as well as the responsible party.

4. In the Forecast computation section, fill in the Forecast horizon field with the number of future points that will be generated by the analysis.

5. In the Data Pool section, specify from where data will be sourced. See how to perform this operation below:

• SE Performance: select this option to use indicators from SoftExpert Performance as data source. To do so, simply click on the button and search for the desired indicator.
• File (*.xls): select this option to use a .xls file as data source. To do so, simply download the available template and insert it filled out using the button.

6. To finish the creation, click on the Save button.

Conclusion

Once the analysis is created, in order for it to be performed, the forecasts must be computed.