hex.glrm

Class GLRMModel

java.lang.Object

water.Iced<T>

water.Keyed<T>

water.Lockable<M>

hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

hex.glrm.GLRMModel

All Implemented Interfaces:

hex.Model.GLRMArchetypes, java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, water.Freezable<GLRMModel>

public class GLRMModel
extends hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>
implements hex.Model.GLRMArchetypes

GLRM (Generalized Low Rank Model). The model seeks to represent an input frame A of dimensions m x n as a product of two smaller matrices X (m x k) and Y (k x n) of rank k each. To this end, the model solves a generic optimization problem Loss(A, XY) + Rx(X) + Ry(Y) -> min_{X,Y} in other words it tries to find X and Y such that XY is close to A (as measured by the loss function), taking into account regularization constraints on X and Y as well. Note that the input frame A may have columns of different types; while output matrices X and Y are always real-valued. The Loss function is assumed to be separable in each element of the matrix, so that Loss(A, XY) = Sum[L_{ij}(A_{ij}, x_i y_j) over i=1..m, j=1..n] the individual loss functions can be different for each element; but in our implementation we assume that L_{ij}'s are constant over rows and may only differ by columns. Thus, L_{ij} == L_j. The regularizers Rx and Ry are assumed to be row-separable: Rx(X) = Sum[rx_i(x_i) for i=1..m] Ry(Y) = Sum[ry_j(x_j) for j=1..n] The output of the model consists of matrices X and Y. There are multiple interpretations of these (see section 5.4 in Boyd's paper). In particular, + The rows of Y (1 x n) can be interpreted as "idealized examples" of input rows (even if rows of Y are always real-valued, while rows of the input data may have any types). Thus, we call them *archetypes* in the code. + The rows of X (1 x k) provide an embedding of each original data row into a lower-dimensional space. Thus, we call them "representations" of the data.

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	GLRMModel.GLRMOutput
static class	GLRMModel.GLRMParameters

Nested classes/interfaces inherited from class hex.Model

hex.Model.BigScore, hex.Model.DeepFeatures, hex.Model.ExemplarMembers, hex.Model.GetMostImportantFeatures, hex.Model.GLRMArchetypes, hex.Model.GridSortBy, hex.Model.InteractionPair, hex.Model.JavaModelStreamWriter, hex.Model.LeafNodeAssignment, hex.Model.Output, hex.Model.Parameters

Field Summary

Fields inherited from class hex.Model

_dist, _output, _parms, _toDelete, _warnings, scoringInfo

Fields inherited from class water.Lockable

_lockers

Fields inherited from class water.Keyed

_key

Constructor Summary

Constructors

Constructor and Description
GLRMModel(water.Key<GLRMModel> selfKey, GLRMModel.GLRMParameters parms, GLRMModel.GLRMOutput output)

Method Summary

Methods

Modifier and Type	Method and Description
GlrmMojoWriter	getMojo()
ModelMetricsGLRM.GlrmModelMetricsBuilder	makeMetricBuilder(java.lang.String[] domain)
protected water.fvec.Frame	predictScoreImpl(water.fvec.Frame orig, water.fvec.Frame adaptedFr, java.lang.String destination_key, water.Job j, boolean computeMetrics)
protected water.Keyed	readAll_impl(water.AutoBuffer ab, water.Futures fs)
protected water.Futures	remove_impl(water.Futures fs)
protected double[]	score0(double[] data, double[] preds)
water.fvec.Frame	scoreArchetypes(water.fvec.Frame frame, water.Key<water.fvec.Frame> destination_key, boolean reverse_transform) Project each archetype into original feature space
ModelMetricsGLRM	scoreMetricsOnly(water.fvec.Frame frame)
water.fvec.Frame	scoreReconstruction(water.fvec.Frame frame, water.Key<water.fvec.Frame> destination_key, boolean reverse_transform)
protected water.AutoBuffer	writeAll_impl(water.AutoBuffer ab)

Methods inherited from class hex.Model

adaptTestForTrain, adaptTestForTrain, addMetrics, addModelMetrics, addWarning, auc, checksum_impl, classification_error, cleanup_adapt, closeBigScorePredict, compareTo, computeDeviances, data, defaultThreshold, deleteCrossValidationModels, deviance, deviance, fillScoringInfo, getDefaultGridSortBy, getToEigenVec, isSupervised, last_scored, lift_top_group, logloss, loss, mae, makeBigScoreTask, makeInteraction, makeInteractions, makeInteractions, makeSchema, makeScoringNames, mean_per_class_error, mse, postProcessPredictions, rmsle, score, score, score, score, score, score0, score0, score0, scoreMetrics, scoring_history, scoringDomains, setupBigScorePredict, testJavaScoring, testJavaScoring, testJavaScoring, toJava, toJava, toJava, toJavaCheckTooBig, toJavaInit, toJavaNCLASSES, toJavaPredictBody, toJavaPROB, toJavaSuper, toString

Methods inherited from class water.Lockable

delete_and_lock, delete_and_lock, delete_and_lock, delete, delete, delete, read_lock, read_lock, read_lock, unlock_all, unlock, unlock, unlock, unlock, update, update, update, write_lock, write_lock, write_lock

Methods inherited from class water.Keyed

checksum, readAll, remove, remove, remove, remove, writeAll

Methods inherited from class water.Iced

asBytes, clone, copyOver, frozenType, read, readExternal, readJSON, reloadFromBytes, toJsonString, write, writeExternal, writeJSON

Methods inherited from class java.lang.Object

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

Constructor Detail

GLRMModel

public GLRMModel(water.Key<GLRMModel> selfKey,
         GLRMModel.GLRMParameters parms,
         GLRMModel.GLRMOutput output)

Method Detail

remove_impl

protected water.Futures remove_impl(water.Futures fs)

Overrides:

remove_impl in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

writeAll_impl

protected water.AutoBuffer writeAll_impl(water.AutoBuffer ab)

Overrides:

writeAll_impl in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

readAll_impl

protected water.Keyed readAll_impl(water.AutoBuffer ab,
                       water.Futures fs)

Overrides:

readAll_impl in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

getMojo

public GlrmMojoWriter getMojo()

Overrides:

getMojo in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

predictScoreImpl

protected water.fvec.Frame predictScoreImpl(water.fvec.Frame orig,
                                water.fvec.Frame adaptedFr,
                                java.lang.String destination_key,
                                water.Job j,
                                boolean computeMetrics)

Overrides:

predictScoreImpl in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

scoreReconstruction

public water.fvec.Frame scoreReconstruction(water.fvec.Frame frame,
                                   water.Key<water.fvec.Frame> destination_key,
                                   boolean reverse_transform)

Specified by:

scoreReconstruction in interface hex.Model.GLRMArchetypes

scoreArchetypes

public water.fvec.Frame scoreArchetypes(water.fvec.Frame frame,
                               water.Key<water.fvec.Frame> destination_key,
                               boolean reverse_transform)

Project each archetype into original feature space

Specified by:

scoreArchetypes in interface hex.Model.GLRMArchetypes

Parameters:

frame - Original training data with m rows and n columns

destination_key - Frame Id for output

Returns:

Frame containing k rows and n columns, where each row corresponds to an archetype

score0

protected double[] score0(double[] data,
              double[] preds)

Specified by:

score0 in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>

scoreMetricsOnly

public ModelMetricsGLRM scoreMetricsOnly(water.fvec.Frame frame)

makeMetricBuilder

public ModelMetricsGLRM.GlrmModelMetricsBuilder makeMetricBuilder(java.lang.String[] domain)

Specified by:

makeMetricBuilder in class hex.Model<GLRMModel,GLRMModel.GLRMParameters,GLRMModel.GLRMOutput>