-
Task
-
Resolution: Fixed
-
Major
-
None
Should allow one to select a handful of attributes to predict a single output attribute. Run selected machine learning method on all subsets/combinations of input attributes and return a ranked list of accuracies in terms of predicting the output attribute. Purpose is to see which attributes have predictive power for the output attribute. Use Kenton's CandE code as a starter, which does what's needed in GWT using Weka.
- is blocking
-
CRI-12 MSC Manual
- Closed
There are no Sub-Tasks for this issue.