| j) Refining your analysis | Understanding the data | |
You can print the resulting tree to use it in reports using the Export option to generate the tree in different formats. To enhance future decision making the tree can be run using Knowledge Builder's inference engine. To enable the tree to be run data fields and attributes need to be mapped. This is done from the Table object of the Training Table by clicking the Auto Map Fields button. You also need to set the inference type between fuzzy or crisp. See Inference Properties in Mined Tree object – Properties tab
The quickest method to run the tree against data is to use the Do command from the Main Agenda of the Knowledge Module (i.e. @Do MinedTree). Knowledge Builder will request values that are necessary to reach a leaf and then the MinedTree object will hold a numeric value which can be displayed using the Debug command (i.e. @Debug MinedTree). The numeric value represents the Probability of the first displayed value of a leaf for a discrete outcome analysis or the Average value for a numeric outcome analysis.