[MRG] Add MAE formula in the regression criteria docs.#8402
[MRG] Add MAE formula in the regression criteria docs.#8402jnothman merged 2 commits intoscikit-learn:masterfrom
Conversation
Location: doc/modules/tree.rst
Codecov Report
@@ Coverage Diff @@
## master #8402 +/- ##
==========================================
+ Coverage 94.75% 94.75% +<.01%
==========================================
Files 342 342
Lines 60813 60816 +3
==========================================
+ Hits 57621 57624 +3
Misses 3192 3192
Continue to review full report at Codecov.
|
| .. math:: | ||
|
|
||
| c_m = \frac{1}{N_m} \sum_{i \in N_m} y_i | ||
|
|
There was a problem hiding this comment.
I think no need to write c_m again, as it already mention above. Better way is to place c_m before Mean Squared Error and Mean Absolute Error once instead of writing in both section individually.
There was a problem hiding this comment.
I think it's fine to repeat the definition of c_m, it makes the definition of MAE. But instead of c_m I would actually use the following notation for that variable name:
\bar{y_m} = \frac{1}{N_m} \sum_{i \in N_m} y_i
ogrisel
left a comment
There was a problem hiding this comment.
Besides the cosmetic comment, +1 on my side.
|
Looks fine, assuming this is correctly the criterion we use. Which I think it is but to be sure: @nelson-liu |
|
yeah, this looks fine to me as well. |
|
Thanks @aashil |
|
And thanks @nelson-liu
…On 24 February 2017 at 13:09, Nelson Liu ***@***.***> wrote:
yeah, this looks fine to me as well.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#8402 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAEz65oKKmG0MDF7a-PjohgKSMQsxMr8ks5rfjvkgaJpZM4MFkdq>
.
|
|
np @jnothman , always happy to weigh in. Disappointed that I haven't had the time to work on scikit-learn recently...hopefully soon. |
Location: doc/modules/tree.rst
Reference Issue
Fixes #8396
What does this implement/fix? Explain your changes.
This PR adds the Mean Absolute Error formula in the regression criteria docs.