Author: isabel
Date: Sun Jan 10 11:15:04 2010
New Revision: 897617
URL: http://svn.apache.org/viewvc?rev=897617&view=rev
Log:
MAHOUT-85 Added implementation for perceptron and winnow trainer.
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearModel.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearTrainer.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainer.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainingMapper.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/TrainingException.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/WinnowTrainer.java
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/LinearModelTest.java
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/PerceptronTrainerTest.java
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/WinnowTrainerTest.java
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearModel.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearModel.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearModel.java
(added)
+++
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearModel.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,103 @@
+/* Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.mahout.classifier.discriminative;
+
+import org.apache.mahout.math.CardinalityException;
+import org.apache.mahout.math.IndexException;
+import org.apache.mahout.math.Vector;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+/**
+ * Classifies a data point using a hyperplane.
+ */
+public class LinearModel {
+ /** Logger for this class. */
+ private static final Logger LOG = LoggerFactory.getLogger(LinearModel.class);
+ /** Represents the direction of the hyperplane found during training.*/
+ private Vector hyperplane;
+ /** Displacement of hyperplane from origin.*/
+ private double bias;
+ /** Classification threshold. */
+ private double threshold;
+
+ /**
+ * Init a linear model with a hyperplane, distance and displacement.
+ * */
+ public LinearModel(final Vector hyperplane, final double displacement, final
double threshold) {
+ this.hyperplane = hyperplane;
+ this.bias = displacement;
+ this.threshold = threshold;
+ }
+
+ /**
+ * Init a linear model with zero displacement and a threshold of 0.5.
+ * */
+ public LinearModel(final Vector hyperplane) {
+ this(hyperplane, 0, 0.5);
+ }
+
+ /**
+ * Classify a point to either belong to the class modeled by this linear
model or not.
+ * @param dataPoint the data point to classify.
+ * @return returns true if data point should be classified as belonging to
this model.
+ * */
+ public boolean classify(final Vector dataPoint) throws CardinalityException,
IndexException {
+ double product = this.hyperplane.dot(dataPoint);
+ if (LOG.isDebugEnabled()) {
+ LOG.debug("model: " + this + " product: " + product + " Bias: " +
this.bias + " threshold: " + this.threshold);
+ }
+ return ((product + this.bias) > this.threshold);
+ }
+
+ /**
+ * Update the hyperplane by adding delta.
+ * @param delta the delta to add to the hyperplane vector.
+ * */
+ public void addDelta(final Vector delta) {
+ this.hyperplane = this.hyperplane.plus(delta);
+ }
+
+ @Override
+ public String toString() {
+ StringBuilder builder = new StringBuilder("Model: ");
+ for (int i = 0; i < this.hyperplane.size(); i++) {
+ builder.append(" ").append(this.hyperplane.get(i));
+ }
+ builder.append(" C: ").append(this.bias);
+ return builder.toString();
+ }
+
+ /**
+ * Shift the bias of the model.
+ * @param factor factor to multiply the bias by.
+ * */
+ public synchronized void shiftBias(double factor) {
+ this.bias = this.bias + factor;
+ }
+
+ /**
+ * Multiply the weight at index by delta.
+ * @param index the index of the element to update.
+ * @param delta the delta to multiply the element with.
+ * */
+ public void timesDelta(int index, double delta) {
+ double element = this.hyperplane.get(index);
+ element = element * delta;
+ this.hyperplane.setQuick(index, element);
+ }
+}
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearTrainer.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearTrainer.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearTrainer.java
(added)
+++
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/LinearTrainer.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,129 @@
+/* Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.mahout.classifier.discriminative;
+
+import org.apache.mahout.math.CardinalityException;
+import org.apache.mahout.math.DenseVector;
+import org.apache.mahout.math.IndexException;
+import org.apache.mahout.math.Matrix;
+import org.apache.mahout.math.Vector;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+/**
+ * Implementors of this class need to provide a way to train linear
+ * discriminative classifiers.
+ *
+ * As this is just the reference implementation we assume that the dataset fits
+ * into main memory - this should be the first thing to change when switching
to
+ * Hadoop.
+ */
+public abstract class LinearTrainer {
+
+ /** Logger for this class. */
+ private static final Logger LOG = LoggerFactory
+ .getLogger(LinearTrainer.class);
+ /** The model to train. */
+ private LinearModel model;
+
+ /**
+ * Initialize the trainer. Distance is initialized to cosine distance, all
+ * weights are represented through a dense vector.
+ *
+ *
+ * @param dimension
+ * number of expected features.
+ * @param threshold
+ * threshold to use for classification.
+ * @param init
+ * initial value of weight vector.
+ * @param initBias
+ * initial classification bias.
+ * */
+ public LinearTrainer(final int dimension, final double threshold,
+ final double init, final double initBias) throws CardinalityException {
+ DenseVector initialWeights = new DenseVector(dimension);
+ initialWeights.assign(init);
+ this.model = new LinearModel(initialWeights, initBias, threshold);
+ }
+
+ /**
+ * Initializes training. Runs through all data points in the training set and
+ * updates the weight vector whenever a classification error occurs.
+ *
+ * Can be called multiple times.
+ *
+ * @param dataset
+ * the dataset to train on. Each column is treated as point.
+ * @param labelset
+ * the set of labels, one for each data point. If the cardinalities
+ * of data- and labelset do not match, a CardinalityException is
+ * thrown
+ * */
+ public void train(final Vector labelset, final Matrix dataset)
+ throws IndexException, CardinalityException, TrainingException {
+ if (labelset.size() != dataset.size()[1]) {
+ throw new CardinalityException();
+ }
+
+ boolean converged = false;
+ int iteration = 0;
+ while (!converged) {
+ if (iteration > 1000)
+ throw new TrainingException(
+ "Too many iterations needed to find hyperplane.");
+
+ converged = true;
+ int columnCount = dataset.size()[1];
+ for (int i = 0; i < columnCount; i++) {
+ Vector dataPoint = dataset.getColumn(i);
+ LOG.debug("Training point: " + dataPoint);
+
+ synchronized (this.model) {
+ boolean prediction = model.classify(dataPoint);
+ double label = labelset.get(i);
+ if ((label <= 0 && prediction) || (label > 0 && !prediction)) {
+ LOG.debug("updating");
+ converged = false;
+ update(label, dataPoint, this.model);
+ }
+ }
+ }
+ }
+ }
+
+ /**
+ * Retrieves the trained model if called after train, otherwise the raw
model.
+ * */
+ public LinearModel getModel() {
+ return this.model;
+ }
+
+ /**
+ * Implement this method to match your training strategy.
+ *
+ * @param model
+ * the model to update.
+ * @param label
+ * the target label of the wrongly classified data point.
+ * @param dataPoint
+ * the data point that was classified incorrectly.
+ * */
+ protected abstract void update(final double label, final Vector dataPoint,
+ LinearModel model);
+
+}
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainer.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainer.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainer.java
(added)
+++
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainer.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,70 @@
+/* Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.mahout.classifier.discriminative;
+
+import org.apache.mahout.math.CardinalityException;
+import org.apache.mahout.math.Vector;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+/**
+ * Implements training accoring to the perceptron update rule.
+ * */
+public class PerceptronTrainer extends LinearTrainer {
+
+ /** Logger for this class. */
+ private static final Logger LOG = LoggerFactory
+ .getLogger(PerceptronTrainer.class);
+ /** Rate the model is to be updated with at each step. */
+ private final double learningRate;
+
+ /**
+ * {...@inheritdoc}
+ *
+ * @param learningRate
+ * rate to update the model with at each step.
+ * */
+ public PerceptronTrainer(int dimension, double threshold,
+ double learningRate, double init, double initBias) throws
CardinalityException {
+ super(dimension, threshold, init, initBias);
+ this.learningRate = learningRate;
+ }
+
+ /**
+ * {...@inheritdoc} Perceptron update works such that in case the predicted
label
+ * does not match the real label, the weight vector is updated as follows: In
+ * case the prediction was positive but should have been negative, the
weight vector
+ * is set to the sum of weight vector and example (multiplied by the
learning rate).
+ *
+ * In case the prediction was negative but should have been positive, the
example
+ * vector (multiplied by the learning rate) is subtracted from the weight
vector.
+ * */
+ @Override
+ protected void update(final double label, final Vector dataPoint,
+ final LinearModel model) {
+ double factor = 1.0;
+ if (label == 0.0)
+ factor = -1.0;
+
+ Vector updateVector = dataPoint.times(factor).times(this.learningRate);
+ LOG.debug("Updatevec: " + updateVector);
+
+ model.addDelta(updateVector);
+ model.shiftBias(factor * this.learningRate);
+ LOG.debug(model.toString());
+ }
+}
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainingMapper.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainingMapper.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainingMapper.java
(added)
+++
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/PerceptronTrainingMapper.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,23 @@
+package org.apache.mahout.classifier.discriminative;
+
+import java.io.IOException;
+
+import org.apache.hadoop.io.Text;
+import org.apache.hadoop.mapreduce.Mapper;
+import org.apache.mahout.math.Vector;
+
+/**
+ * Mapper for perceptron training. Strategy for parallelization:
+ * 1) Train separate models on training data samples. Each training
+ * data sample must fit into main memory.
+ * 2) Average all trained models into one.
+ * */
+public class PerceptronTrainingMapper extends
+ Mapper<Boolean, Vector, Text, LinearModel> {
+
+ @Override
+ protected void map(final Boolean key, final Vector value, Context context)
throws IOException, InterruptedException {
+
+ }
+
+}
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/TrainingException.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/TrainingException.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/TrainingException.java
(added)
+++
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/TrainingException.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,18 @@
+package org.apache.mahout.classifier.discriminative;
+
+/**
+ * This exception is thrown in case training fails. E.g. training with an
algorithm
+ * that can find linear separating hyperplanes only on a training set that is
not
+ * linearly separable.
+ * */
+public class TrainingException extends Exception {
+ /** Serialization id. */
+ private static final long serialVersionUID = 388611231310145397L;
+
+ /**
+ * Init with message string describing the cause of the exception.
+ * */
+ public TrainingException(final String message) {
+ super(message);
+ }
+}
Added:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/WinnowTrainer.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/WinnowTrainer.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/WinnowTrainer.java
(added)
+++
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/classifier/discriminative/WinnowTrainer.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,90 @@
+/* Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.mahout.classifier.discriminative;
+
+import java.util.Iterator;
+
+import org.apache.mahout.math.CardinalityException;
+import org.apache.mahout.math.Vector;
+import org.apache.mahout.math.Vector.Element;
+
+/**
+ * This class implements training according to the winnow update algorithm.
+ */
+public class WinnowTrainer extends LinearTrainer {
+
+ /** Promotion step to multiply weights with on update. */
+ private final double promotionStep;
+
+ public WinnowTrainer(final int dimension, final double promotionStep,
+ final double threshold, final double init, final double initBias) {
+ super(dimension, threshold, init, initBias);
+ this.promotionStep = promotionStep;
+ }
+
+ /** {...@inheritdoc} */
+ public WinnowTrainer(final int dimension, final double promotionStep)
+ throws CardinalityException {
+ this(dimension, promotionStep, 0.5, 1, 0);
+ }
+
+ /**
+ * Initializes with dimension and promotionStep of 2.
+ *
+ * @param dimension
+ * number of features.
+ * */
+ public WinnowTrainer(final int dimension) {
+ this(dimension, 2);
+ }
+
+ /**
+ * {...@inheritdoc} Winnow update works such that in case the predicted label
+ * does not match the real label, the weight vector is updated as follows: In
+ * case the prediction was positiv but should have been negative, all entries
+ * in the weight vector that correspond to non null features in the example
+ * are doubled.
+ *
+ * In case the prediction was negative but should have been positive, all
+ * entries in the weight vector that correspond to non null features in the
+ * example are halfed.
+ * */
+ @Override
+ protected void update(final double label, final Vector dataPoint,
+ LinearModel model) {
+ if (label > 0) {
+ // case one
+ Vector updateVector = dataPoint.times(1 / this.promotionStep);
+ System.out.println("Winnow update positive: " + updateVector);
+ Iterator<Element> iter = updateVector.iterateNonZero();
+ while (iter.hasNext()) {
+ Element element = iter.next();
+ model.timesDelta(element.index(), element.get());
+ }
+ } else {
+ // case two
+ Vector updateVector = dataPoint.times(1 / this.promotionStep);
+ System.out.println("Winnow update negative: " + updateVector);
+ Iterator<Element> iter = updateVector.iterateNonZero();
+ while (iter.hasNext()) {
+ Element element = iter.next();
+ model.timesDelta(element.index(), element.get());
+ }
+ }
+ System.out.println(model);
+ }
+}
Added:
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/LinearModelTest.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/LinearModelTest.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/LinearModelTest.java
(added)
+++
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/LinearModelTest.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,61 @@
+package org.apache.mahout.classifier.discriminative;
+
+import org.apache.mahout.math.DenseVector;
+import org.apache.mahout.math.Vector;
+
+import junit.framework.TestCase;
+
+public class LinearModelTest extends TestCase {
+
+ private LinearModel model;
+ private Vector hyperplane;
+
+ protected void setUp() throws Exception {
+ super.setUp();
+ double[] values = {0.0, 1.0, 0.0, 1.0, 0.0};
+ this.hyperplane = new DenseVector(values);
+ this.model = new LinearModel(this.hyperplane, 0.1, 0.5);
+ }
+
+ public void testClassify() {
+ double[] valuesFalse = {1.0, 0.0, 1.0, 0.0, 1.0};
+ Vector dataPointFalse = new DenseVector(valuesFalse);
+ assertFalse(this.model.classify(dataPointFalse));
+
+ double[] valuesTrue = {0.0, 1.0, 0.0, 1.0, 0.0};
+ Vector dataPointTrue = new DenseVector(valuesTrue);
+ assertTrue(this.model.classify(dataPointTrue));
+ }
+
+ public void testAddDelta() {
+ double[] values = {1.0, -1.0, 1.0, -1.0, 1.0};
+ this.model.addDelta(new DenseVector(values));
+
+ double[] valuesFalse = {1.0, 0.0, 1.0, 0.0, 1.0};
+ Vector dataPointFalse = new DenseVector(valuesFalse);
+ assertTrue(this.model.classify(dataPointFalse));
+
+ double[] valuesTrue = {0.0, 1.0, 0.0, 1.0, 0.0};
+ Vector dataPointTrue = new DenseVector(valuesTrue);
+ assertFalse(this.model.classify(dataPointTrue));
+ }
+
+ public void testTimesDelta() {
+ double[] values = {-1.0, -1.0, -1.0, -1.0, -1.0};
+ this.model.addDelta(new DenseVector(values));
+ double[] dotval = {-1.0, -1.0, -1.0, -1.0, -1.0};
+
+ for (int i = 0; i < dotval.length; i++) {
+ this.model.timesDelta(i, dotval[i]);
+ }
+
+ double[] valuesFalse = {1.0, 0.0, 1.0, 0.0, 1.0};
+ Vector dataPointFalse = new DenseVector(valuesFalse);
+ assertTrue(this.model.classify(dataPointFalse));
+
+ double[] valuesTrue = {0.0, 1.0, 0.0, 1.0, 0.0};
+ Vector dataPointTrue = new DenseVector(valuesTrue);
+ assertFalse(this.model.classify(dataPointTrue));
+ }
+
+}
Added:
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/PerceptronTrainerTest.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/PerceptronTrainerTest.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/PerceptronTrainerTest.java
(added)
+++
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/PerceptronTrainerTest.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,40 @@
+package org.apache.mahout.classifier.discriminative;
+
+import org.apache.mahout.math.DenseMatrix;
+import org.apache.mahout.math.DenseVector;
+import org.apache.mahout.math.Matrix;
+import org.apache.mahout.math.Vector;
+
+import junit.framework.TestCase;
+
+public class PerceptronTrainerTest extends TestCase {
+
+ private PerceptronTrainer trainer;
+
+ protected void setUp() throws Exception {
+ super.setUp();
+ trainer = new PerceptronTrainer(3, 0.5, 0.1, 1.0, 1.0);
+ }
+
+ public void testUpdate() throws TrainingException {
+ double[] labels = { 1.0, 1.0, 1.0, 0.0 };
+ Vector labelset = new DenseVector(labels);
+ double[][] values = new double[3][4];
+ for (int i = 0; i < 3; i++) {
+ values[i][0] = 1.0;
+ values[i][1] = 1.0;
+ values[i][2] = 1.0;
+ values[i][3] = 1.0;
+ }
+ values[1][0] = 0.0;
+ values[2][0] = 0.0;
+ values[1][1] = 0.0;
+ values[2][2] = 0.0;
+
+ Matrix dataset = new DenseMatrix(values);
+ this.trainer.train(labelset, dataset);
+ assertFalse(this.trainer.getModel().classify(dataset.getColumn(3)));
+ assertTrue(this.trainer.getModel().classify(dataset.getColumn(0)));
+ }
+
+}
Added:
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/WinnowTrainerTest.java
URL:
http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/WinnowTrainerTest.java?rev=897617&view=auto
==============================================================================
---
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/WinnowTrainerTest.java
(added)
+++
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/classifier/discriminative/WinnowTrainerTest.java
Sun Jan 10 11:15:04 2010
@@ -0,0 +1,41 @@
+package org.apache.mahout.classifier.discriminative;
+
+import org.apache.mahout.math.DenseMatrix;
+import org.apache.mahout.math.DenseVector;
+import org.apache.mahout.math.Matrix;
+import org.apache.mahout.math.Vector;
+
+import junit.framework.TestCase;
+
+
+public class WinnowTrainerTest extends TestCase {
+
+ private WinnowTrainer trainer;
+
+ protected void setUp() throws Exception {
+ super.setUp();
+ trainer = new WinnowTrainer(3);
+ }
+
+ public void testUpdate() throws Exception {
+ double[] labels = { 0.0, 0.0, 0.0, 1.0 };
+ Vector labelset = new DenseVector(labels);
+ double[][] values = new double[3][4];
+ for (int i = 0; i < 3; i++) {
+ values[i][0] = 1.0;
+ values[i][1] = 1.0;
+ values[i][2] = 1.0;
+ values[i][3] = 1.0;
+ }
+ values[1][0] = 0.0;
+ values[2][0] = 0.0;
+ values[1][1] = 0.0;
+ values[2][2] = 0.0;
+
+ Matrix dataset = new DenseMatrix(values);
+ trainer.train(labelset, dataset);
+ assertTrue(trainer.getModel().classify(dataset.getColumn(3)));
+ assertFalse(trainer.getModel().classify(dataset.getColumn(0)));
+ }
+
+}
