[
https://issues.apache.org/jira/browse/SPARK-3181?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15384138#comment-15384138
]
Yanbo Liang edited comment on SPARK-3181 at 7/19/16 1:26 PM:
-------------------------------------------------------------
[~dbtsai] Should the {{HuberRegression}} share codebase with
{{LinearRegression}} in ML? I saw you suggested to do that. But I found it's
hard to share codebase between the two regression methods although they
accomplish the same goal. Since the output of {{fit}} will be different:
{{LinearRegressionModel}} and {{HuberRegressionModel}}. The former only
contains {{coefficients, intercept}}, but the latter contains {{coefficients,
intercept, scale/sigma}}. It will also involve save/load compatibility issue if
we combine the two models become one. One trick method is we can drop
{{scale/sigma}} and make the {{fit}} by Huber cost function still output
{{LinearRegressionModel}}, but I don't think it's an appropriate way. So I
would like to implement {{HuberRegression}} in a separate file and looking
forward to hear your thought. Thanks!
was (Author: yanboliang):
[~dbtsai] Should the {{HuberRegression}} share codebase with
{{LinearRegression}} in ML? I saw you suggested to do that. But I found it's
hard to share codebase between the two regression methods although they
accomplish the same goal. Since the output of {{fit}} will be different:
{{LinearRegressionModel}} and {{HuberRegressionModel}}. The former only
contains {{coefficients, intercept}}, but the latter contains {{coefficients,
intercept, scale/sigma}}. It will also involve save/load compatibility issue if
we combine the two model become one. One trick method is we can drop
{{scale/sigma}} and make the {{fit}} by Huber cost function still output
{{LinearRegressionModel}}, but I don't think it's an appropriate way. So I
would like to implement {{HuberRegression}} in a separate file and looking
forward to hear your thought. Thanks!
> Add Robust Regression Algorithm with Huber Estimator
> ----------------------------------------------------
>
> Key: SPARK-3181
> URL: https://issues.apache.org/jira/browse/SPARK-3181
> Project: Spark
> Issue Type: New Feature
> Components: ML, MLlib
> Reporter: Fan Jiang
> Assignee: Yanbo Liang
> Labels: features
> Original Estimate: 0h
> Remaining Estimate: 0h
>
> Linear least square estimates assume the error has normal distribution and
> can behave badly when the errors are heavy-tailed. In practical we get
> various types of data. We need to include Robust Regression to employ a
> fitting criterion that is not as vulnerable as least square.
> In 1973, Huber introduced M-estimation for regression which stands for
> "maximum likelihood type". The method is resistant to outliers in the
> response variable and has been widely used.
> The new feature for MLlib will contain 3 new files
> /main/scala/org/apache/spark/mllib/regression/RobustRegression.scala
> /test/scala/org/apache/spark/mllib/regression/RobustRegressionSuite.scala
> /main/scala/org/apache/spark/examples/mllib/HuberRobustRegression.scala
> and one new class HuberRobustGradient in
> /main/scala/org/apache/spark/mllib/optimization/Gradient.scala
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
