search

salesforce / TransmogrifAI

TransmogrifAI (pronounced trăns-mŏgˈrə-fī) is an AutoML library for building modular, reusable, strongly typed machine learning workflows on Apache Spark with minimal hand-tuning
https://transmogrif.ai
BSD 3-Clause "New" or "Revised" License
2.24k stars 393 forks source link

why does not linear svc model contain "probability"s in the output of "score()" function？ #396

Closed liuxiaodong008008 closed 5 years ago

liuxiaodong008008 commented 5 years ago

I found linear svc model does not contain probabilitys in the output of score() function. And if linear svc does not have probabilitys, how do TransmogrifAI calculate AuRoC in the summary()?

tovbinm commented 5 years ago

The probability is not provided by the underlying LinearSVCModel Spark model and therefore is not produced.

When computing metrics such as aupr, auroc, error, precision, recall, f1, only the label, prediction and raw prediction values are being used. E.g. - https://github.com/salesforce/TransmogrifAI/blob/master/core/src/main/scala/com/salesforce/op/evaluators/OpBinaryClassificationEvaluator.scala#L126)