Github user yanboliang commented on a diff in the pull request:

    https://github.com/apache/spark/pull/15394#discussion_r83001469
  
    --- Diff: 
mllib/src/main/scala/org/apache/spark/ml/optim/WeightedLeastSquares.scala ---
    @@ -45,34 +47,48 @@ private[ml] class WeightedLeastSquaresModel(
      * formulation:
      *
      * min,,x,z,, 1/2 sum,,i,, w,,i,, (a,,i,,^T^ x + z - b,,i,,)^2^ / sum,,i,, 
w_i
    - *   + 1/2 lambda / delta sum,,j,, (sigma,,j,, x,,j,,)^2^,
    + *   + lambda / delta (1/2 (1 - alpha) sum,,j,, (sigma,,j,, x,,j,,)^2^
    + *   + alpha sum,,j,, abs(sigma,,j,, x,,j,,)),
      *
    - * where lambda is the regularization parameter, and delta and sigma,,j,, 
are controlled by
    - * [[standardizeLabel]] and [[standardizeFeatures]], respectively.
    + * where lambda is the regularization parameter, alpha is the ElasticNet 
mixing parameter,
    + * and delta and sigma,,j,, are controlled by [[standardizeLabel]] and 
[[standardizeFeatures]],
    + * respectively.
    --- End diff --
    
    Should we document that we always solve the normal equations in the 
standardized space and convert back to the original space for output?


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

Reply via email to