Log Class
Some of the trainers accept a loss parameter that will be used for training. It is also known as loss function, objective function, or optimization score function.
- Inheritance
-
builtins.objectLog
Constructor
Log()
Examples
###############################################################################
# Log Loss
from nimbusml.linear_model import FastLinearBinaryClassifier
from nimbusml.loss import Log
# specifying the loss function as a string keyword
trainer1 = FastLinearBinaryClassifier(loss='log')
# can also use loss class instead of string
trainer1 = FastLinearBinaryClassifier(loss=Log()) # equivalent to loss='log'
Remarks
Losses can be specified either as a string or a loss object. When loss is specified as one of these strings, the default values are used for the loss parameters. To change the default parameters, a loss object should be used, as seen in examples below.
Each trainer supports only a subset of the losses mentioned above. To get the supported losses and the default loss, please refer to the documentation page for the specific trainer.
The Log loss
for classification. Its string name is 'log'
.
It can be used for AveragedPerceptronBinaryClassifier, FastLinearBinaryClassifier, FastLinearClassifier, SgdBinaryClassifier.