Package org.tribuo.classification.sgd.fm
Class FMClassificationOptions
java.lang.Object
org.tribuo.classification.sgd.fm.FMClassificationOptions
- All Implemented Interfaces:
com.oracle.labs.mlrg.olcut.config.Options
,ClassificationOptions<FMClassificationTrainer>
public class FMClassificationOptions
extends Object
implements ClassificationOptions<FMClassificationTrainer>
CLI options for training a factorization machine classifier.
-
Nested Class Summary
Modifier and TypeClassDescriptionstatic enum
Available loss types. -
Field Summary
Modifier and TypeFieldDescriptionint
Number of SGD epochs.int
Factor size.int
Log the objective after n examples.int
Minibatch size.Loss function.double
Variance of the initialization gaussian.The gradient descent algorithm options.Fields inherited from interface com.oracle.labs.mlrg.olcut.config.Options
header
-
Constructor Summary
-
Method Summary
Modifier and TypeMethodDescriptiongetLoss()
Returns the loss function specified in the arguments.Constructs the trainer based on the provided arguments.Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
Methods inherited from interface com.oracle.labs.mlrg.olcut.config.Options
getOptionsDescription
-
Field Details
-
sgoOptions
The gradient descent algorithm options. -
fmEpochs
@Option(longName="fm-epochs", usage="Number of SGD epochs.") public int fmEpochsNumber of SGD epochs. -
fmObjective
@Option(longName="fm-objective", usage="Loss function.") public FMClassificationOptions.LossEnum fmObjectiveLoss function. -
fmLoggingInterval
@Option(longName="fm-logging-interval", usage="Log the objective after <int> examples.") public int fmLoggingIntervalLog the objective after n examples. -
fmMinibatchSize
@Option(longName="fm-minibatch-size", usage="Minibatch size.") public int fmMinibatchSizeMinibatch size. -
fmFactorSize
@Option(longName="fm-factor-size", usage="Factor size.") public int fmFactorSizeFactor size. -
fmVariance
@Option(longName="fm-variance", usage="Variance of the initialization gaussian.") public double fmVarianceVariance of the initialization gaussian.
-
-
Constructor Details
-
FMClassificationOptions
public FMClassificationOptions()
-
-
Method Details
-
getLoss
Returns the loss function specified in the arguments.- Returns:
- The loss function.
-
getTrainer
Description copied from interface:ClassificationOptions
Constructs the trainer based on the provided arguments.- Specified by:
getTrainer
in interfaceClassificationOptions<FMClassificationTrainer>
- Returns:
- The trainer.
-