Examples of DenseDoubleMatrix

de.jungblut.math.dense.DenseDoubleMatrix
org.apache.hama.commons.math.DenseDoubleMatrix
Dense double matrix implementation, internally uses two dimensional double arrays.
org.apache.hama.ml.math.DenseDoubleMatrix
Dense double matrix implementation, internally uses two dimensional double arrays.

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

    double[][] mat = { { 1, 2, 3 }, { 4, 5, 6 }, { 7, 8, 9 } };
    double[][] mat2 = { { 10, 20 }, { 30, 40 }, { 50, 60 } };
    double[][][] mats = { mat, mat2 };


    DenseDoubleMatrix[] matrices = new DenseDoubleMatrix[] {
        new DenseDoubleMatrix(mat), new DenseDoubleMatrix(mat2) };


    SmallMLPMessage message = new SmallMLPMessage(owner, true, matrices);


    Configuration conf = new Configuration();
    String strPath = "/tmp/testSmallMLPMessage";

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

        { 0.7, 0.8, 0.9 } };
    double[][] prevMat2 = { { 1, 2 }, { 3, 4 }, { 5, 6 } };
    double[][][] prevMats = { prevMat, prevMat2 };


    DenseDoubleMatrix[] matrices = new DenseDoubleMatrix[] {
        new DenseDoubleMatrix(mat), new DenseDoubleMatrix(mat2) };


    DenseDoubleMatrix[] prevMatrices = new DenseDoubleMatrix[] {
        new DenseDoubleMatrix(prevMat), new DenseDoubleMatrix(prevMat2) };


    boolean terminated = false;
    SmallMLPMessage message = new SmallMLPMessage(owner, terminated, matrices,
        prevMatrices);

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

    DoubleMatrix[] weightUpdates = new DoubleMatrix[this.inMemoryModel.weightMatrixList
        .size()];
    for (int i = 0; i < weightUpdates.length; ++i) {
      int row = this.inMemoryModel.weightMatrixList.get(i).getRowCount();
      int col = this.inMemoryModel.weightMatrixList.get(i).getColumnCount();
      weightUpdates[i] = new DenseDoubleMatrix(row, col);
    }


    // continue to train
    double avgTrainingError = 0.0;
    LongWritable key = new LongWritable();

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

      int sizePrevLayer = this.layerSizeList.get(layerIdx - 1);
      // row count equals to size of current size and column count equals to
      // size of previous layer
      int row = isFinalLayer ? size : size - 1;
      int col = sizePrevLayer;
      DoubleMatrix weightMatrix = new DenseDoubleMatrix(row, col);
      // initialize weights
      final Random rnd = new Random();
      weightMatrix.applyToElements(new DoubleFunction() {
        @Override
        public double apply(double value) {
          return rnd.nextDouble() - 0.5;
        }


        @Override
        public double applyDerivative(double value) {
          throw new UnsupportedOperationException("");
        }
      });
      this.weightMatrixList.add(weightMatrix);
      this.prevWeightUpdatesList.add(new DenseDoubleMatrix(row, col));
      this.squashingFunctionList.add(squashingFunction);
    }
    return layerIdx;
  }

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

    this.weightMatrixList = new ArrayList<DoubleMatrix>();
    this.prevWeightUpdatesList = new ArrayList<DoubleMatrix>();
    for (int i = 0; i < numOfMatrices; ++i) {
      DoubleMatrix matrix = MatrixWritable.read(input);
      this.weightMatrixList.add(matrix);
      this.prevWeightUpdatesList.add(new DenseDoubleMatrix(
          matrix.getRowCount(), matrix.getColumnCount()));
    }


  }

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

    DoubleVector output = internalResults.get(internalResults.size() - 1);
    // initialize weight update matrices
    DenseDoubleMatrix[] weightUpdateMatrices = new DenseDoubleMatrix[this.weightMatrixList
        .size()];
    for (int m = 0; m < weightUpdateMatrices.length; ++m) {
      weightUpdateMatrices[m] = new DenseDoubleMatrix(this.weightMatrixList
          .get(m).getRowCount(), this.weightMatrixList.get(m).getColumnCount());
    }
    DoubleVector deltaVec = new DenseDoubleVector(
        this.layerSizeList.get(this.layerSizeList.size() - 1));

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

  private void initializeWeightMatrix() {
    this.weightMatrice = new DenseDoubleMatrix[this.numberOfLayers - 1];
    // each layer contains one bias neuron
    for (int i = 0; i < this.numberOfLayers - 1; ++i) {
      // add weights for bias
      this.weightMatrice[i] = new DenseDoubleMatrix(this.layerSizeArray[i] + 1,
          this.layerSizeArray[i + 1]);


      this.weightMatrice[i].applyToElements(new DoubleFunction() {


        private Random rnd = new Random();

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

  private void initializePrevWeightUpdateMatrix() {
    this.prevWeightUpdateMatrices = new DenseDoubleMatrix[this.numberOfLayers - 1];
    for (int i = 0; i < this.prevWeightUpdateMatrices.length; ++i) {
      int row = this.layerSizeArray[i] + 1;
      int col = this.layerSizeArray[i + 1];
      this.prevWeightUpdateMatrices[i] = new DenseDoubleMatrix(row, col);
    }
  }

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

  DenseDoubleMatrix[] trainByInstance(DoubleVector trainingInstance)
      throws Exception {
    // initialize weight update matrices
    DenseDoubleMatrix[] weightUpdateMatrices = new DenseDoubleMatrix[this.layerSizeArray.length - 1];
    for (int m = 0; m < weightUpdateMatrices.length; ++m) {
      weightUpdateMatrices[m] = new DenseDoubleMatrix(
          this.layerSizeArray[m] + 1, this.layerSizeArray[m + 1]);
    }


    if (trainingInstance == null) {
      return weightUpdateMatrices;
    }


    double[] trainingVec = trainingInstance.toArray();
    double[] trainingFeature = Arrays.copyOfRange(trainingVec, 0,
        this.layerSizeArray[0]);
    double[] trainingLabels = Arrays.copyOfRange(trainingVec,
        this.layerSizeArray[0], trainingVec.length);


    DoubleVector trainingFeatureVec = new DenseDoubleVector(trainingFeature);
    List<double[]> outputCache = this.outputInternal(trainingFeatureVec);


    // calculate the delta of output layer
    double[] delta = new double[this.layerSizeArray[this.layerSizeArray.length - 1]];
    double[] outputLayerOutput = outputCache.get(outputCache.size() - 1);
    double[] lastHiddenLayerOutput = outputCache.get(outputCache.size() - 2);


    DenseDoubleMatrix prevWeightUpdateMatrix = this.prevWeightUpdateMatrices[this.prevWeightUpdateMatrices.length - 1];
    for (int j = 0; j < delta.length; ++j) {
      delta[j] = this.costFunction.applyDerivative(trainingLabels[j],
          outputLayerOutput[j]);
      // add regularization term
      if (this.regularization != 0.0) {
        double derivativeRegularization = 0.0;
        DenseDoubleMatrix weightMatrix = this.weightMatrice[this.weightMatrice.length - 1];
        for (int k = 0; k < this.layerSizeArray[this.layerSizeArray.length - 1]; ++k) {
          derivativeRegularization += weightMatrix.get(k, j);
        }
        derivativeRegularization /= this.layerSizeArray[this.layerSizeArray.length - 1];
        delta[j] += this.regularization * derivativeRegularization;
      }

View Full Code Here

Examples of org.apache.hama.ml.math.DenseDoubleMatrix

    int prevLayerIdx = curLayerIdx - 1;
    double[] delta = new double[this.layerSizeArray[curLayerIdx]];
    double[] curLayerOutput = outputCache.get(curLayerIdx);
    double[] prevLayerOutput = outputCache.get(prevLayerIdx);


    DenseDoubleMatrix prevWeightUpdateMatrix = this.prevWeightUpdateMatrices[curLayerIdx - 1];
    // for each neuron j in nextLayer, calculate the delta
    for (int j = 0; j < delta.length; ++j) {
      // aggregate delta from next layer
      for (int k = 0; k < nextLayerDelta.length; ++k) {
        double weight = this.weightMatrice[curLayerIdx].get(j, k);

View Full Code Here

0 1 2 3 4 5

TOP

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.