I am a student who uses Watson analytics for my data analytic course. WA is a great tool, and I am happy to use it. If WA provides more tools for data refinement, it will be great. I found a minor bug (if I am wrong, please guide me the right way). When I create the predictive model, WA gives me wonderful results. When I create the predictive model for the categorical target, I checked the decision tree. On the right-hand side of the decision tree, WA provides three options: 1) Full tree, 2) Top 5 nodes, and 3) Bottom 5 nodes. When I select 1, the result is fine. When I select 2, the result is fine too. But when I select 3 (Bottom 5 nodes), labeling bottom 5 nodes on the decision diagram does not change. Labeling 5 nodes in the decision diagram are the same labeling for the top 5 nodes although I turn to bottom 5 nodes in my option. Bottom 5 nodes in the list box do not match in the tree diagram because tree diagram keeps still the top 5 nodes.
Why is it useful?
|Who would benefit from this IDEA?||If the bug was fixed, it will benefit to all users|
How should it work?