hex

Class NeuralNet

java.lang.Object

water.Iced

water.api.Constants

water.api.RequestStatics

water.api.RequestArguments

water.api.RequestQueries

water.api.RequestBuilders

water.api.Request

water.Request2

water.Func

water.Job

water.Job.FrameJob

water.Job.ColumnsJob

water.Job.ColumnsResJob

water.Job.ModelJobWithoutClassificationField

water.Job.ModelJob

water.Job.ValidatedJob

hex.NeuralNet

All Implemented Interfaces:

java.lang.Cloneable, Freezable

public class NeuralNet
extends Job.ValidatedJob

Neural network.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	NeuralNet.Activation Activation functions
static class	NeuralNet.Errors
static class	NeuralNet.ExecutionMode
static class	NeuralNet.InitialWeightDistribution
static class	NeuralNet.Loss Loss functions CrossEntropy is recommended
static class	NeuralNet.NeuralNetModel
static class	NeuralNet.NeuralNetScore

Nested classes/interfaces inherited from class water.Job.ValidatedJob

Job.ValidatedJob.Response2CMAdaptor

Nested classes/interfaces inherited from class water.Job

Job.ChunkProgress, Job.ChunkProgressJob, Job.ColumnsJob, Job.ColumnsResJob, Job.Fail, Job.FrameJob, Job.HexJob, Job.JobCancelledException, Job.JobState, Job.List, Job.ModelJob, Job.ModelJobWithoutClassificationField, Job.Progress, Job.ProgressMonitor, Job.ValidatedJob

Nested classes/interfaces inherited from class water.Request2

Request2.ColumnSelect, Request2.Dependent, Request2.DoClassBoolean, Request2.DRFCopyDataBoolean, Request2.MultiVecSelect, Request2.MultiVecSelectType, Request2.TypeaheadKey, Request2.VecClassSelect, Request2.VecSelect

Nested classes/interfaces inherited from class water.api.Request

Request.API, Request.Default, Request.Filter, Request.Validator<V>

Nested classes/interfaces inherited from class water.api.RequestBuilders

RequestBuilders.ArrayBuilder, RequestBuilders.ArrayHeaderRowBuilder, RequestBuilders.ArrayRowBuilder, RequestBuilders.ArrayRowElementBuilder, RequestBuilders.ArrayRowSingleColBuilder, RequestBuilders.BooleanStringBuilder, RequestBuilders.Builder, RequestBuilders.ElementBuilder, RequestBuilders.HideBuilder, RequestBuilders.KeyCellBuilder, RequestBuilders.KeyElementBuilder, RequestBuilders.KeyLinkElementBuilder, RequestBuilders.KeyMinAvgMaxBuilder, RequestBuilders.NoCaptionObjectBuilder, RequestBuilders.ObjectBuilder, RequestBuilders.PaginatedTable, RequestBuilders.PreFormattedBuilder, RequestBuilders.Response, RequestBuilders.ResponseInfo, RequestBuilders.WarningCellBuilder

Nested classes/interfaces inherited from class water.api.RequestArguments

RequestArguments.Argument<T>, RequestArguments.Bool, RequestArguments.CaseModeSelect, RequestArguments.CaseSelect, RequestArguments.ClassifyBool, RequestArguments.DRFCopyDataBool, RequestArguments.EnumArgument<T extends java.lang.Enum<T>>, RequestArguments.ExistingFile, RequestArguments.FrameClassVec, RequestArguments.FrameKeyMultiVec, RequestArguments.FrameKeyVec, RequestArguments.GeneralFile, RequestArguments.H2OCategoryStrata, RequestArguments.H2OCategoryWeights, RequestArguments.H2OExistingKey, RequestArguments.H2OGLMModelKey, RequestArguments.H2OHexKey, RequestArguments.H2OHexKeyCol, RequestArguments.H2OIllegalArgumentException, RequestArguments.H2OKey, RequestArguments.H2OKey2, RequestArguments.H2OKMeansModelKey, RequestArguments.H2OModelKey<TM extends OldModel,TK extends TypeaheadKeysRequest>, RequestArguments.HexAllColumnSelect, RequestArguments.HexColumnSelect, RequestArguments.HexKeyClassCol, RequestArguments.HexNonClassColumnSelect, RequestArguments.HexNonConstantColumnSelect, RequestArguments.HexPCAColumnSelect, RequestArguments.InputCheckBox, RequestArguments.InputSelect<T>, RequestArguments.InputText<T>, RequestArguments.Int, RequestArguments.LongInt, RequestArguments.MultipleSelect<T>, RequestArguments.MultipleText<T>, RequestArguments.NTree, RequestArguments.NumberSequence, RequestArguments.NumberSequenceFloat, RequestArguments.Real, RequestArguments.Record<T>, RequestArguments.RFModelKey, RequestArguments.RSeq, RequestArguments.RSeqFloat, RequestArguments.Str, RequestArguments.StringList, RequestArguments.TypeaheadInputText<T>

Nested classes/interfaces inherited from class water.api.RequestStatics

RequestStatics.RequestType

Nested classes/interfaces inherited from class water.api.Constants

Constants.Extensions, Constants.Schemes, Constants.Suffixes

Field Summary

Fields

Modifier and Type	Field and Description
NeuralNet.Activation	activation
boolean	diagnostics
static DocGen.FieldDoc[]	DOC_FIELDS
static java.lang.String	DOC_GET
double	epochs
boolean	expert_mode
boolean	fast_mode
int[]	hidden
NeuralNet.InitialWeightDistribution	initial_weight_distribution
double	initial_weight_scale
double	input_dropout_ratio
double	l1
double	l2
NeuralNet.Loss	loss
double	max_w2
NeuralNet.ExecutionMode	mode
long	momentum_ramp
double	momentum_stable
double	momentum_start
double	rate
double	rate_annealing
double	rate_decay
static boolean	running
double	score_interval
long	score_training
long	score_validation
long	seed
long	warmup_samples

Fields inherited from class water.Job.ValidatedJob

_cmDomain, _names, _responseName, _sourceResponseDomain, _train, _valid, _validResponse, _validResponseDomain, validation

Fields inherited from class water.Job.ModelJob

classification

Fields inherited from class water.Job.ColumnsResJob

response

Fields inherited from class water.Job.ColumnsJob

cols, ignored_cols, ignored_cols_by_name

Fields inherited from class water.Job.FrameJob

source

Fields inherited from class water.Job

_fjtask, description, destination_key, end_time, exception, job_key, LIST, start_time, state

Fields inherited from class water.Request2

_parms, response_info

Fields inherited from class water.api.Request

_requestHelp, SUPPORTS_ONLY_V1, SUPPORTS_ONLY_V2, SUPPORTS_V1_V2

Fields inherited from class water.api.RequestBuilders

ARRAY_BUILDER, ARRAY_HEADER_ROW_BUILDER, ARRAY_ROW_BUILDER, ARRAY_ROW_ELEMENT_BUILDER, ARRAY_ROW_SINGLECOL_BUILDER, ELEMENT_BUILDER, GSON_BUILDER, OBJECT_BUILDER, ROOT_OBJECT

Fields inherited from class water.api.RequestQueries

_queryHtml

Fields inherited from class water.api.RequestArguments

_arguments

Fields inherited from class water.api.Constants

ALPHA, ARGUMENTS, AUC, BASE, BEST_THRESHOLD, BETA_EPS, BIN_LIMIT, BROWSE, BUCKET, BUILT_IN_KEY_JOBS, CANCELLED, CARDINALITY, CASE, CASE_MODE, CHUNK, CLASS, CLOUD_HEALTH, CLOUD_NAME, CLOUD_SIZE, CLOUD_UPTIME_MILLIS, CLUSTERS, COEFFICIENTS, COL_INDEX, COLS, COLUMN_NAME, COLUMNS_DISPLAY, CONSENSUS, CONTENTS, COUNT, DATA_KEY, DEPTH, DESCRIPTION, DEST_KEY, DTHRESHOLDS, ELAPSED, END_TIME, ENUM_DOMAIN_SIZE, ERROR, ESCAPE_NAN, EXCLUSIVE_SPLIT_LIMIT, EXPRESSION, FAILED, FAMILY, FEATURES, FILE, FILES, FILTER, FIRST_CHUNK, FJ_QUEUE_HI, FJ_QUEUE_LO, FJ_THREADS_HI, FJ_THREADS_LO, FREE_DISK, FREE_MEM, HEADER, HEIGHT, HELP, IGNORE, ITEMS, ITERATIVE_CM, JOB, JOB_KEY, JOBS, JSON_H2O, KEY, KEYS, LAMBDA, LAST_CONTACT, LIMIT, LINK, LOCKED, MAX, MAX_DISK, MAX_ITER, MAX_MEM, MAX_ROWS, MEAN, MIN, MODEL_KEY, MODELS, MORE, MTRY, MTRY_NODES, NAME, NEG_X, NO_CM, NODE, NODE_HEALTH, NODE_NAME, NODES, NORMALIZE, NUM_COLS, NUM_CPUS, NUM_FAILED, NUM_KEYS, NUM_MISSING_VALUES, NUM_ROWS, NUM_SUCCEEDED, NUM_TREES, OBJECT, OFFSET, OOBEE, PARALLEL, PARSER_TYPE, PATH, PREVIEW, PREVIOUS_MODEL_KEY, PRIOR, PROGRESS, PROGRESS_KEY, PROGRESS_TOTAL, REDIRECT, REDIRECT_ARGS, REPLICATION_FACTOR, REQUEST_TIME, RESPONSE, RHO, ROW, ROW_SIZE, ROWS, RPCS, SAMPLE, SAMPLING_STRATEGY, SCALE, SEED, SENT_ROWS, SEPARATOR, SIZE, SOURCE_KEY, STACK_TRACES, START_TIME, STAT_TYPE, STATUS, STEP, STRATA_SAMPLES, SUCCEEDED, SYSTEM_LOAD, TASK_KEY, TCPS_ACTIVE, TCPS_DUTY, TIME, TO_ENUM, TOT_MEM, TREE_COUNT, TREE_DEPTH, TREE_LEAVES, TREE_NUM, TREES, TWEEDIE_POWER, TYPE, URL, USE_NON_LOCAL_DATA, VALUE, VALUE_SIZE, VALUE_TYPE, VARIANCE, VERSION, VIEW, WARNINGS, WEIGHT, WEIGHTS, WIDTH, X, XVAL, Y

Constructor Summary

Constructors

Constructor and Description
NeuralNet()

Method Summary

Methods

Modifier and Type	Method and Description
static NeuralNet.Errors	eval(Layer[] ls, long n, long[][] cm)
static NeuralNet.Errors	eval(Layer[] ls, Vec[] vecs, Vec resp, long n, long[][] cm)
void	execImpl() The real implementation which should be provided by ancestors.
static java.lang.String	link(Key k, java.lang.String content)
float	progress() Return progress of this job.
protected void	queryArgumentValueSet(RequestArguments.Argument arg, java.util.Properties inputArgs) Overwrite this method to be able to change / disable values of other arguments on certain argument changes.
static void	reChunk(Vec[] vecs) Makes sure small datasets are spread over enough chunks to parallelize training.
protected RequestBuilders.Response	redirect()
protected void	registered(RequestServer.API_VERSION ver) Iterates over fields and their annotations, and creates argument handlers.
java.lang.String	speedDescription() Description of a speed criteria: msecs/frob
long	speedValue() Value of the described speed criteria: msecs/frob
boolean	toHTML(java.lang.StringBuilder sb)

Methods inherited from class water.Job.ValidatedJob

getCMDomain, getOrigValidation, getValidAdaptor, getValidation, getVectorDomain, hasValidation, init, prepareValidationWithModel, toJSON

Methods inherited from class water.Job.ColumnsJob

selectFrame, selectVecs

Methods inherited from class water.Job

all, cancel, cancel, cancel, checkIdx, defaultDestKey, defaultJobKey, dest, findJob, findJobByDest, fork, get, getState, gridParallelism, invoke, isCancelledOrCrashed, isCancelledXX, isCrashed, isDone, isEnded, isRunning, isRunning, onCancelled, remove, runTimeMs, self, serve, start, waitUntilJobEnded, waitUntilJobEnded

Methods inherited from class water.Func

cleanup, emptyLTrash, exec, gtrash, gtrash, ltrash, ltrash

Methods inherited from class water.Request2

create, fillResponseInfo, filterNaCols, input, logStart, makeJsonBox, serveGrid, set, split, superServeGrid, supportedVersions, toString

Methods inherited from class water.api.Request

addToNavbar, addToNavbar, addToNavbar, DocExampleFail, DocExampleSucc, href, href, hrefType, HTMLHelp, htmlTemplate, initializeNavBar, log, mapTypeahead, ReSTHelp, serve, serveJava, serveResponse, toDocGET, toJava, wrap, wrap, wrap, writeJSONFields

Methods inherited from class water.api.RequestBuilders

build, buildJSONResponseBox, buildResponseHeader, name

Methods inherited from class water.api.RequestQueries

buildQuery, checkArguments

Methods inherited from class water.api.RequestArguments

arguments, argumentsToJson, frameColumnNameToIndex, vaCategoryNames, vaCategoryNames, vaColumnNameToIndex

Methods inherited from class water.api.RequestStatics

checkJsonName, encodeRedirectArgs, JSON2HTML, jsonError, requestName, Str2JSON

Methods inherited from class water.Iced

clone, frozenType, init, newInstance, read, toDocField, write, writeJSON

Methods inherited from class java.lang.Object

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

Field Detail

DOC_FIELDS

public static DocGen.FieldDoc[] DOC_FIELDS

DOC_GET

public static final java.lang.String DOC_GET

See Also:

Constant Field Values

mode

@Request.API(help="Execution Mode",
             filter=Request.Default.class,
             json=true)
public NeuralNet.ExecutionMode mode

activation

@Request.API(help="Activation function",
             filter=Request.Default.class,
             json=true)
public NeuralNet.Activation activation

input_dropout_ratio

@Request.API(help="Input layer dropout ratio",
             filter=Request.Default.class,
             dmin=0.0,
             dmax=1.0,
             json=true)
public double input_dropout_ratio

hidden

@Request.API(help="Hidden layer sizes, e.g. 1000, 1000. Grid search: (100, 100), (200, 200)",
             filter=Request.Default.class,
             json=true)
public int[] hidden

rate

@Request.API(help="Learning rate (higher => less stable, lower => slower convergence)",
             filter=Request.Default.class,
             dmin=0.0,
             dmax=1.0,
             json=true)
public double rate

rate_annealing

@Request.API(help="Learning rate annealing: rate / (1 + rate_annealing * samples)",
             filter=Request.Default.class,
             dmin=0.0,
             dmax=1.0,
             json=true)
public double rate_annealing

l1

@Request.API(help="L1 regularization, can add stability",
             filter=Request.Default.class,
             dmin=0.0,
             dmax=1.0,
             json=true)
public double l1

l2

@Request.API(help="L2 regularization, can add stability",
             filter=Request.Default.class,
             dmin=0.0,
             dmax=1.0,
             json=true)
public double l2

momentum_start

@Request.API(help="Initial momentum at the beginning of training",
             filter=Request.Default.class,
             dmin=0.0,
             json=true)
public double momentum_start

momentum_ramp

@Request.API(help="Number of training samples for which momentum increases",
             filter=Request.Default.class,
             lmin=0L,
             json=true)
public long momentum_ramp

momentum_stable

@Request.API(help="Final momentum after the ramp is over",
             filter=Request.Default.class,
             dmin=0.0,
             json=true)
public double momentum_stable

epochs

@Request.API(help="How many times the dataset should be iterated (streamed), can be less than 1.0",
             filter=Request.Default.class,
             dmin=0.0,
             json=true)
public double epochs

seed

@Request.API(help="Seed for random numbers (reproducible results for single-threaded only, cf. Hogwild)",
             filter=Request.Default.class,
             json=true)
public long seed

expert_mode

@Request.API(help="Enable expert mode",
             filter=Request.Default.class,
             json=true)
public boolean expert_mode

initial_weight_distribution

@Request.API(help="Initial Weight Distribution",
             filter=Request.Default.class,
             json=true)
public NeuralNet.InitialWeightDistribution initial_weight_distribution

initial_weight_scale

@Request.API(help="Uniform: -value...value, Normal: stddev)",
             filter=Request.Default.class,
             dmin=0.0,
             json=true)
public double initial_weight_scale

loss

@Request.API(help="Loss function",
             filter=Request.Default.class,
             json=true)
public NeuralNet.Loss loss

rate_decay

@Request.API(help="Learning rate decay factor between layers (N-th layer: rate*alpha^(N-1))",
             filter=Request.Default.class,
             dmin=0.0,
             json=true)
public double rate_decay

max_w2

@Request.API(help="Constraint for squared sum of incoming weights per unit",
             filter=Request.Default.class,
             json=true)
public double max_w2

warmup_samples

@Request.API(help="Number of samples to train with non-distributed mode for improved stability",
             filter=Request.Default.class,
             lmin=0L,
             json=true)
public long warmup_samples

score_training

@Request.API(help="Number of training set samples for scoring (0 for all)",
             filter=Request.Default.class,
             lmin=0L,
             json=true)
public long score_training

score_validation

@Request.API(help="Number of validation set samples for scoring (0 for all)",
             filter=Request.Default.class,
             lmin=0L,
             json=true)
public long score_validation

score_interval

@Request.API(help="Minimum interval (in seconds) between scoring",
             filter=Request.Default.class,
             dmin=0.0,
             json=true)
public double score_interval

diagnostics

@Request.API(help="Enable diagnostics for hidden layers",
             filter=Request.Default.class,
             json=true)
public boolean diagnostics

fast_mode

@Request.API(help="Enable fast mode (minor approximation in back-propagation)",
             filter=Request.Default.class,
             json=true)
public boolean fast_mode

running

public static volatile boolean running

Constructor Detail

NeuralNet

public NeuralNet()

Method Detail

toHTML

public boolean toHTML(java.lang.StringBuilder sb)

Overrides:

toHTML in class Request

registered

protected void registered(RequestServer.API_VERSION ver)

Description copied from class: Request2

Iterates over fields and their annotations, and creates argument handlers.

Overrides:

registered in class Job.ColumnsResJob

queryArgumentValueSet

protected void queryArgumentValueSet(RequestArguments.Argument arg,
                         java.util.Properties inputArgs)

Description copied from class: RequestQueries

Overwrite this method to be able to change / disable values of other arguments on certain argument changes. This is done for both query checking and request checking.

Overrides:

queryArgumentValueSet in class RequestQueries

execImpl

public final void execImpl()

Description copied from class: Func

The real implementation which should be provided by ancestors.

Overrides:

execImpl in class Func

progress

public float progress()

Description copied from class: Job

Return progress of this job.

Overrides:

progress in class Job

Returns:

the value in interval <0,1> representing job progress.

eval

public static NeuralNet.Errors eval(Layer[] ls,
                    Vec[] vecs,
                    Vec resp,
                    long n,
                    long[][] cm)

eval

public static NeuralNet.Errors eval(Layer[] ls,
                    long n,
                    long[][] cm)

redirect

protected RequestBuilders.Response redirect()

Overrides:

redirect in class Job

link

public static java.lang.String link(Key k,
                    java.lang.String content)

speedDescription

public java.lang.String speedDescription()

Description copied from class: Job

Description of a speed criteria: msecs/frob

Overrides:

speedDescription in class Job

speedValue

public long speedValue()

Description copied from class: Job

Value of the described speed criteria: msecs/frob

Overrides:

speedValue in class Job

reChunk

public static void reChunk(Vec[] vecs)

Makes sure small datasets are spread over enough chunks to parallelize training.