hex.deeplearning

Class DeepLearning

java.lang.Object

water.Iced<T>

water.Keyed

water.Job<M>

hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

hex.deeplearning.DeepLearning

All Implemented Interfaces:

java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, water.Freezable

public class DeepLearning
extends hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

Deep Learning Neural Net implementation based on MRTask

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
class	DeepLearning.DeepLearningDriver

Nested classes/interfaces inherited from class hex.ModelBuilder

hex.ModelBuilder.BuilderVisibility

Nested classes/interfaces inherited from class water.Job

water.Job.JobCancelledException, water.Job.JobState, water.Job.Progress, water.Job.ProgressUpdate, water.Job.ValidationMessage

Nested classes/interfaces inherited from class water.Keyed

water.Keyed.BinarySerializer<X extends water.Keyed>

Field Summary

Fields inherited from class hex.ModelBuilder

_fold, _nclass, _offset, _parms, _response, _train, _valid, _vresponse, _weights

Fields inherited from class water.Job

_description, _dest, _end_time, _exception, _messages, _progressKey, _start_time, _state, LIST

Fields inherited from class water.Keyed

_key, EMPTY_KEY_LIST

Constructor Summary

Constructors

Constructor and Description
DeepLearning(DeepLearningParameters parms) Main constructor from Deep Learning parameters

Method Summary

Methods

Modifier and Type	Method and Description
hex.ModelBuilder.BuilderVisibility	builderVisibility()
hex.ModelCategory[]	can_build() Types of models we can build with DeepLearning
protected void	checkMemoryFootPrint()
void	init(boolean expensive) Initialize the ModelBuilder, validating all arguments and preparing the training frame.
boolean	isSupervised()
void	modifyParmsForCrossValidationMainModel(int N)
void	modifyParmsForCrossValidationSplits(int i, int N, water.Key<hex.Model> model_id)
long	progressUnits()
hex.schemas.ModelBuilderSchema	schema()
water.Job<DeepLearningModel>	trainModelImpl(long work, boolean restartTimer) Start the DeepLearning training Job on an F/J thread.

Methods inherited from class hex.ModelBuilder

canBeDone, cancel, checkDistributions, clearInitState, computeCrossValidation, computePriorClassDistribution, createModelBuilder, deleteProgressKey, error_count, getAlgo, getAlgo, getAlgo, getAlgoFullName, getModelBuilder, getModelBuilders, getModelClass, hasFoldCol, hasOffsetCol, hasWeightCol, ignoreBadColumns, ignoreStringColumns, isClassifier, nclasses, nFoldCV, numSpecialCols, registerModelBuilder, response, responseMean, separateFeatureVecs, train, trainModel, updateModelOutput, valid, vresponse

Methods inherited from class water.Job

block, cancel, checksum_impl, clearValidationErrors, createProgressKey, dest, done, done, error_count_or_uninitialized, error, failed, get, hide, info, isCancelledOrCrashed, isDone, isRunning, isRunning, isStopped, jobKey, jobs, message, msec, progress_msg, progress, remove_impl, start, update, update, update, update, updateValidationMessages, validationErrors, warn

Methods inherited from class water.Keyed

checksum, getBinarySerializer, getPublishedKeys, remove, remove, remove, remove

Methods inherited from class water.Iced

clone, frozenType, read_impl, read, readExternal, readJSON_impl, readJSON, toJsonString, write_impl, write, writeExternal, writeJSON_impl, writeJSON

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DeepLearning

public DeepLearning(DeepLearningParameters parms)

Main constructor from Deep Learning parameters

Parameters:

parms -

Method Detail

can_build

public hex.ModelCategory[] can_build()

Types of models we can build with DeepLearning

Specified by:

can_build in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

Returns:

schema

public hex.schemas.ModelBuilderSchema schema()

Specified by:

schema in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

builderVisibility

public hex.ModelBuilder.BuilderVisibility builderVisibility()

Specified by:

builderVisibility in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

isSupervised

public boolean isSupervised()

Overrides:

isSupervised in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

trainModelImpl

public water.Job<DeepLearningModel> trainModelImpl(long work,
                                          boolean restartTimer)

Start the DeepLearning training Job on an F/J thread.

Specified by:

trainModelImpl in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

Parameters:

work -

restartTimer -

progressUnits

public long progressUnits()

Specified by:

progressUnits in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

init

public void init(boolean expensive)

Initialize the ModelBuilder, validating all arguments and preparing the training frame. This call is expected to be overridden in the subclasses and each subclass will start with "super.init();". This call is made by the front-end whenever the GUI is clicked, and needs to be fast; heavy-weight prep needs to wait for the trainModel() call. Validate the very large number of arguments in the DL Parameter directly.

Overrides:

init in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

checkMemoryFootPrint

protected void checkMemoryFootPrint()

Overrides:

checkMemoryFootPrint in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

modifyParmsForCrossValidationSplits

public void modifyParmsForCrossValidationSplits(int i,
                                       int N,
                                       water.Key<hex.Model> model_id)

Overrides:

modifyParmsForCrossValidationSplits in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>

modifyParmsForCrossValidationMainModel

public void modifyParmsForCrossValidationMainModel(int N)

Overrides:

modifyParmsForCrossValidationMainModel in class hex.ModelBuilder<DeepLearningModel,DeepLearningParameters,DeepLearningModel.DeepLearningModelOutput>