ModelSelection MaxRSweep replacement bug

[exalate-issue-sync]

No description provided.

[exalate-issue-sync]

Wendy Wong commented: It looks like the replacement step in H2O-3 modelselection maxrsweep stops too early:

!image-20220707-174319.png|width=50%!

The replacement step should use replace C101 with C89 with an increase of R2.

[exalate-issue-sync]

Wendy Wong commented: The issue here is this:

After forward step, say I have selected ABC as my predictors.

Then, I go into the replacement step and aiming to try out _BC where I will pick the subcarrier in first position. Say, I chose DBC as the best.

The next replacement should be D_C where I will pick the subcarrier in the middle position. However, I implemented it wrong and was starting with A_C. I fixed this and compared my result with customer result and they match.

[h2o-ops]

JIRA Issue Details

Jira Issue: PUBDEV-8758
Assignee: Wendy Wong
Reporter: Wendy Wong
State: Resolved
Fix Version: 3.38.0.1
Attachments: Available (Count: 1)
Development PRs: Available

[h2o-ops]

Attachments From Jira

Attachment Name: image-20220707-174319.png
Attached By: Wendy Wong
File Link:https://h2o-3-jira-github-migration.s3.amazonaws.com/PUBDEV-8758/image-20220707-174319.png

[h2o-ops]

Linked PRs from JIRA

#6246
https://github.com/h2oai/h2oai-serving/pull/1079