spark-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Congrui Yi <>
Subject MLlib-Missing Regularization Parameter and Intercept for Logistic Regression
Date Fri, 13 Jun 2014 18:38:25 GMT
Hi All,
I'm new to Spark and currently exploring around the ML package mllib. Two
questions about LogisticRegressionWithSGD in MLlib classification package.
I have checked the source code on github and found that the class
LogisticRegressionWithSGD has "regParam" as its private argument. I cannot
access it directly since it is private. Also, the 4 "train" methods in the
companion object all uses regParam = 0.0. Why aren't users allowed to use
nonzero regParam like in SVM? Is it because the feature has not been
implemented yet?

As shown in the documentation, the class extends from
GeneralizedLinearAlgorithm which has a method "setIntercept" but again this
is not an argument in the train methods of the companion object
LogisticRegressionWithSGD. Shouldn't the users be allowed to choose whether
they want an intercept since the feature is already there?

Thank you!


Congrui Yi

View this message in context:
Sent from the Apache Spark User List mailing list archive at

View raw message