Predictive Model Tutorial - Run Analysis

Each model includes a sample data set (OfficeStar) that can be found under the Tutorials dropdown on the Enginius Dashboard.

The remainder of the tutorials use the Predictive OfficeStar data set as the starting data set.

Predictive Model Run Analysis Setting (OfficeStar Default Settings)

Predictive modeling requires:

Target Variable

• Select the type of choice data that you have available.

Calibration data

• Calibration data: Specify the data block that contains your predictive variables.
• Target variable: Select the column name from the above calibration data block that contains your target variable.
  • Box-Cox transforms the predictors: See Appendix for full description.
  • Log transform the target variable (available when using a Continuous or Discrete-continuous target variable):  The log transformation can be used to make highly skewed distributions less skewed. This can be valuable for making patterns in the data more interpretable.
• Cross –validation: See Appendix for full description.

Out-of-sample predictions

• Apply predictive modeling to out-of-sample data: allows you assess the predictive validity of the model by using the results from the calibration with an additional data block.
• Out-of-sample data (available when "Apply predictive modeling" is checked): Specify the data block that contains the out-of-sample data. This data should contain the same variables as the calibration data with the exception that there should be no target variable column.
  Example: Calibration data:
  Out-of-sample data:
  

Your Enginius report can be generated in many different formats. Clicking the globe beside the Run button will allow you to select a new report format.

After selecting the desired model options, click  to generate the report in the desired format.