Nobuyuki Kuromatsu created SPARK-7085:
-----------------------------------------
Summary: Inconsistent default miniBatchFraction parameters in the
train methods of RidgeRegression
Key: SPARK-7085
URL: https://issues.apache.org/jira/browse/SPARK-7085
Project: Spark
Issue Type: Bug
Components: MLlib
Affects Versions: 1.3.1
Reporter: Nobuyuki Kuromatsu
Priority: Minor
The miniBatchFraction parameter in the train method called with 4 arguments is
0.01, that is,
{code:title=RidgeRegression.scala|borderStyle=solid}
def train(
input: RDD[LabeledPoint],
numIterations: Int,
stepSize: Double,
regParam: Double): RidgeRegressionModel = {
train(input, numIterations, stepSize, regParam, 0.01)
}
{code}
but, the parameter is 1.0 in the other train methods. For example,
{code:title=RidgeRegression.scala|borderStyle=solid}
def train(
input: RDD[LabeledPoint],
numIterations: Int): RidgeRegressionModel = {
train(input, numIterations, 1.0, 0.01, 1.0)
}
{code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
