Discretisation Options- Manual

[General Discretise Options]  [Auto Discretise Options]

Discretisation is performed using a supplied list of split points. Each bin is assigned a number similar to the auto methods. Manual discretisation is commonly used to apply splits to test data similar to that performed by auto methods on the training data.

Unlike auto discretisation methods, manual discretisation can only be applied to a single field at a time. This is because the split points provided will be specific to the range of values within the selected field. To apply a set of split points to more than one field, use the clone feature to duplicate the node and set the fieldname option accordingly in each new node.

 

Field

Selects the single field to be manually discretised using the supplied list of split points.

 

Split points

Enter a new split point that will be added to the list. The list of split points that will be used to discretise the selected field is displayed below the new split point field. You must enter at least one split point (in order to produce two bins).

 

Add

Click this button to add a new value into the list of split points.

 

Delete

Click this button to delete the currently selected value from list of split points.

 

Start at bin

The number used to indicate the first bin. By default, numbering starts at 0 and subsequent bins are numbered sequentially. You can use the spin control to increase or decrease the start bin value.

 

Convert discretised field(s) to categorical

After the discretised data has been generated, this option converts the field to categorical values. This procedure allows the field to be used as the target field for feature selection, or mining activities.