To use the classifier the vector parameters must be defined. The parameters are defined by adding toolbox results in the Classification pane.

The following operations are available:

• Add - Activates Classification Parameter Setup
• Edit - Change an existing parameter
• Up - moves a parameter up
• Down - moves a parameter down
• Delete - removes a parameter
• Active - activates / deactivates a parameter
• Required - sets parameters to be required for the classifier to return true
• Always included - will be included also when undefined or not calculated
• Parameter weight - change parameter weight
• Normalize - normalizes the calculated weight coeffisients

Weights

• Automatic weight calculation - activates automatic weight calculation method
  • 1 / ( max-min) | 1 / std | 1 / 3*std | none
• A - calculates weights

Constraints

• Maximum distance - defined the maximum distance to be accepted
• Minimum separation - required separation between classes to be accepted
• Minimum quality - the percentage of defined results in the classification vector

Note: if no constraints are active the closest reference will be selected as the best class

Automatic weights optimize the distance measure based on the reference values

Classification Parameter Setup

• Active - enable parameter
• Required for classification - sets parameters to be required for the classifier to return true
• Always included in distance - will be included also when undefined or not calculated
• Name - name of parameter
• Parameter - toolbox result name
• Parameter weight - user defined weight
  • Note: in addition to the normalized weights
• Conversion - activates result conversion
  • Square, Square Root and Cube Root
  • Note: An area can be converted by taking the Square Root
• Statistics - displays the reference statistics for this parameter