This is the second screen of the three screen rule extraction wizard.
The Back<< button moves you backward to field selection>
The Next>> button moves you forward to rule extraction

Row Selection... Training Set

• Choose what rows you wish to use for extracting the rules.
• You can drag the black numbers, or double click a black number to over-type. HINT: Remember to press Enter after you over-type a black number.

Row Selection... Testing Set

• Choose what rows you wish to use for testing the performance of the extracted rules.
• You can drag the black numbers, or double click a black number to over-type. HINT: Remember to press Enter after you over-type a black number.

Row Selection... Hints

• For data mining by anomaly detection, use the same rows for both training and testing.
• For model building, select the Testing Set so none of these rows are seen during the training of the rule extraction.
• For very large data sets, you will get faster performance if you avoid selecting too much data in either set.
• In order for the training set and testing set to be of the same character, it is important that data be randomly shuffled.