Description
GradientDescent$.runMiniBatchSGD currently has two overloads, one of which uses a default convergence tolerance which is not documented.
It should either use a default argument (rather than overloading) or specify the default convergence tolerance in the method overload's documentation.
Attachments
Issue Links
- is related to
-
SPARK-9660 ML 1.5 QA: API: New Scala APIs, docs
- Resolved
- links to