Fixes for quantile regression #148
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Out of convenience, MLJLM defines residuals as$X\theta - y$ , however the literature typically does the opposite. This matters only if the loss is asymmetrical which is the case for the pinball loss used in the quantile regression. The report in #147 exposed this.
Note that the report in #147 also compared the performance of an IP method with the result obtained by MLJLM. The objective achieved using MLJLM is about 9% worse than that from the IP method. When using a non-IP method from the quantreg package, the objective function is 1-2% better. This is not worrying per se, in the case presented LBFGS just stops at a non-optimal point, likely gets stuck in a very flat zone.