[ML] Fixes incorrect feature importance visualization for Data Frame Analytics classification#150816
Conversation
qn895
added
release_note:fix
:ml
Feature:Data Frame Analytics
|
Pinging @elastic/ml-ui (:ml) |
| const [currentClass, setCurrentClass] = useState<string>( | ||
| getStringBasedClassName(topClasses[0].class_name) | ||
| ); | ||
| const selectedClass = topClasses.find( |
There was a problem hiding this comment.
The description of the fix refers to True and False base classes, but is this change also the cause of the change in behavior here with the bank_classification_1 MLQA bootstrap job where before I was seeing values greater than 1 for yes and no classes:
Before:
With this fix:
There was a problem hiding this comment.
AFAICS, the fix makes total sense here: for the second row, the model predicts the class "no" with the probability of 0.986. This means that the prediction probability of "yes" is 0.014. This is precisely what the screenshot above shows now 🚀
The screenshot before this fix shows that the decision graph briefly exceeds 1.0 in for prediction probability, which is nonsense.
peteharverson
left a comment
peteharverson
left a comment
There was a problem hiding this comment.
Tested and LGTM.
When testing this, looks like there is another unrelated issue where the popover only populates on page 1 of the results. On any other page it is empty for me:
Thanks for catching that. Looks like this was a regression that was introduced in #125023. I've added a fix as well as functional tests to check that the popover shows up correctly after table pagination is updated. |
|
@elasticmachine merge upstream |
|
@elasticmachine merge upstream |
|
@elasticmachine merge upstream |
💚 Build Succeeded
Metrics [docs]Async chunks
History
To update your PR or re-run it, just comment with: cc @qn895 |
…Analytics classification (elastic#150816) Co-authored-by: Kibana Machine <42973632+kibanamachine@users.noreply.github.com> (cherry picked from commit c2476d2)
💚 All backports created successfully
Note: Successful backport PRs will be merged automatically after passing CI. Questions ?Please refer to the Backport tool documentation |
…Frame Analytics classification (#150816) (#151094) # Backport This will backport the following commits from `main` to `8.7`: - [[ML] Fixes incorrect feature importance visualization for Data Frame Analytics classification (#150816)](#150816) <!--- Backport version: 8.9.7 --> ### Questions ? Please refer to the [Backport tool documentation](https://github.com/sqren/backport) <!--BACKPORT [{"author":{"name":"Quynh Nguyen (Quinn)","email":"43350163+qn895@users.noreply.github.com"},"sourceCommit":{"committedDate":"2023-02-14T02:43:46Z","message":"[ML] Fixes incorrect feature importance visualization for Data Frame Analytics classification (#150816)\n\nCo-authored-by: Kibana Machine <42973632+kibanamachine@users.noreply.github.com>","sha":"c2476d240e5a5a979af215057bb7f2bd40b9f6fe","branchLabelMapping":{"^v8.8.0$":"main","^v(\\d+).(\\d+).\\d+$":"$1.$2"}},"sourcePullRequest":{"labels":["release_note:fix",":ml","Feature:Data Frame Analytics","v8.7.0","v8.8.0"],"number":150816,"url":"https://github.com/elastic/kibana/pull/150816","mergeCommit":{"message":"[ML] Fixes incorrect feature importance visualization for Data Frame Analytics classification (#150816)\n\nCo-authored-by: Kibana Machine <42973632+kibanamachine@users.noreply.github.com>","sha":"c2476d240e5a5a979af215057bb7f2bd40b9f6fe"}},"sourceBranch":"main","suggestedTargetBranches":["8.7"],"targetPullRequestStates":[{"branch":"8.7","label":"v8.7.0","labelRegex":"^v(\\d+).(\\d+).\\d+$","isSourceBranch":false,"state":"NOT_CREATED"},{"branch":"main","label":"v8.8.0","labelRegex":"^v8.8.0$","isSourceBranch":true,"state":"MERGED","url":"https://github.com/elastic/kibana/pull/150816","number":150816,"mergeCommit":{"message":"[ML] Fixes incorrect feature importance visualization for Data Frame Analytics classification (#150816)\n\nCo-authored-by: Kibana Machine <42973632+kibanamachine@users.noreply.github.com>","sha":"c2476d240e5a5a979af215057bb7f2bd40b9f6fe"}}]}] BACKPORT--> Co-authored-by: Quynh Nguyen (Quinn) <43350163+qn895@users.noreply.github.com>
7 participants
Summary
This PR fixes #146122 where the decision graph could briefly exceed 1.0 in for prediction probability. The issue was with the 'true' or 'false' string values not matching correctly with title-cased 'True'/'False'.
Before:
After:
Checklist