Difference between revisions of "KCM Cross Validation Values"
(→Notes/Comments) |
(→Notes/Comments) |
||
| Line 10: | Line 10: | ||
** Only on KCMs with mapping type of Step-to-KC | ** Only on KCMs with mapping type of Step-to-KC | ||
** Need to store two numbers, the cross validation number (RMSE: decimal between 0-1) and the number of rows. | ** Need to store two numbers, the cross validation number (RMSE: decimal between 0-1) and the number of rows. | ||
| + | ** The input for computing Cross Validation RMSE is SSSS formatted file | ||
* There are also a few changes to the current datashop interface | * There are also a few changes to the current datashop interface | ||
Revision as of 20:00, 12 May 2010
Status: Prioritization Needed
User Story
As a [type of user], I want to [perform some task] so that I can [achieve some goal].
Notes/Comments
- I would like to provide DataShop with an algorithm to calculate cross validation values for each KC Model. -- Hui Cheng, 5/10/2010
- Only on KCMs with mapping type of Step-to-KC
- Need to store two numbers, the cross validation number (RMSE: decimal between 0-1) and the number of rows.
- The input for computing Cross Validation RMSE is SSSS formatted file
- There are also a few changes to the current datashop interface
- Add number of parameters for BIC an AIC
- Change the way opportunity numbers are counted. This will affect Learning Curve--LFA Values.
See complete DataShop Feature Wish List.
