I agree, implementing StreamingRidgeRegressionWithSGD and StreamingLassoWithSGD following StreamingLinearRegressionWithSGD is make sense.
2014-12-02 6:00 GMT+08:00 Joanne Contact <joannenetw...@gmail.com>: > Hi Gurus, > > I did not look at the code yet. I wonder if > StreamingLinearRegressionWithSGD > <http://spark.apache.org/docs/latest/api/java/org/apache/spark/mllib/regression/StreamingLinearRegressionWithSGD.html> > > is equivalent to > LinearRegressionWithSGD > <http://spark.apache.org/docs/latest/api/java/org/apache/spark/mllib/regression/LinearRegressionWithSGD.html>with > starting weights of the current batch as the ending weights of the last > batch? > > Since RidgeRegressionModel > <http://spark.apache.org/docs/latest/api/java/org/apache/spark/mllib/regression/RidgeRegressionModel.html> > does > not seem to have a streaming version, just wonder if this way will suffice. > > > Thanks! > > J >