[ https://issues.apache.org/jira/browse/SPARK-37285?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17442177#comment-17442177 ]
Apache Spark commented on SPARK-37285: -------------------------------------- User 'taosiyuan163' has created a pull request for this issue: https://github.com/apache/spark/pull/34553 > Add Weight of Evidence and Information value to ml.feature > ---------------------------------------------------------- > > Key: SPARK-37285 > URL: https://issues.apache.org/jira/browse/SPARK-37285 > Project: Spark > Issue Type: New Feature > Components: ML > Affects Versions: 3.2.0 > Reporter: Simon Tao > Priority: Major > > The weight of evidence (WOE) and information value (IV) provide a great > framework for exploratory analysis and variable screening for binary > classifiers as well as beneficial and help us analyze multiple points as > listed below: > 1. Helps check the linear relationship of a feature with its dependent > feature to be used in the model. > 2. Is a good variable transformation method for both continuous and > categorical features. > 3. Is better than on-hot encoding as this method of variable transformation > does not increase the complexity of the model. > 4. Detect linear and non-linear relationships. > 5. Be useful in feature selection. > 6. Is a good measure of the predictive power of a feature and it also helps > point out the suspicious feature. -- This message was sent by Atlassian Jira (v8.20.1#820001) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org