Hi, I've implemented a class with measures for evaluation of multiclass classification (as well as unit tests). They are per class and averaged Precision, Recall and F1-measure. As far as I know, in Spark, there is binary classification evaluator only, given that Spark's Bayesian classifier supports multiclass. I've submitted a pull request https://github.com/apache/spark/pull/1155 following the guidelines on https://cwiki.apache.org/confluence/display/SPARK/Contributing+to+Spark Admins didn't yet verify my patch. I have few questions: 1)Do I need to contact somebody to be verified (my Github profile isn't linked with my personal page)? 2)Spark users, would you like other measures to be implemented or any other features for multiclass evaluator, for example, Accuracy or returning the confusion matrix?
Best regards, Alexander
