lookupservice.java

一个数据挖掘软件ALPHAMINERR的整个过程的JAVA版源代码

      sap[4].setValue("1");
      sap[4].setDescription("discretization level");
      sap[4].setStatus(0);

      sap[5].setName("attributeLevels");
      sap[5].setType("int[]");
      sap[5].setValue("null");
      sap[5].setDescription("discretization levels of attributes if anisotropic discretization");
      sap[5].setStatus(0);

      sap[6].setName("lambda");
      sap[6].setType("double");
      sap[6].setValue("1");
      sap[6].setDescription("value of regularization parameter");
      sap[6].setStatus(0);

      sapVec.addElement(sap);
    }
    if ( settings instanceof SupportVectorSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 8;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sap[0].setName("svmType");
      sap[0].setType("int");
      sap[0].setValue("0");
      sap[0].setDescription("type of SCM (classification, regression, nu-SVM, ...");
      sap[0].setStatus(0);

      sap[1].setName("kernelType");
      sap[1].setType("int");
      sap[1].setValue("2");
      sap[1].setDescription("kernel type (linear, poly, RBF, sigmoid)");
      sap[1].setStatus(0);

      sap[2].setName("degree");
      sap[2].setType("double");
      sap[2].setValue("3");
      sap[2].setDescription("polynomial degree");
      sap[2].setStatus(0);

      sap[3].setName("gamma");
      sap[3].setType("double");
      sap[3].setValue("1");
      sap[3].setDescription("exponential coefficient");
      sap[3].setStatus(0);

      sap[4].setName("coef0");
      sap[4].setType("double");
      sap[4].setValue("0");
      sap[4].setDescription("absolute term");
      sap[4].setStatus(0);

      sap[5].setName("lossEpsilon");
      sap[5].setType("double");
      sap[5].setValue("0.1");
      sap[5].setDescription("loss epsilon in regularization");
      sap[5].setStatus(0);

      sap[6].setName("C");
      sap[6].setType("double");
      sap[6].setValue("1");
      sap[6].setDescription("value of inverse regularization parameter");
      sap[6].setStatus(0);

      sap[7].setName("nu");
      sap[7].setType("double");
      sap[7].setValue("0.5");
      sap[7].setDescription("nu in nu-SVM");
      sap[7].setStatus(0);

      sapVec.addElement(sap);
    }
    if ( settings instanceof NeuralNetworkSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 6;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sap[0].setName("learningType");
      sap[0].setType("int");
      sap[0].setValue("0");
      sap[0].setDescription("Type of learning algorithm (backpropagation, with momentum)");
      sap[0].setStatus(0);

      sap[1].setName("autoBuildNetwork");
      sap[1].setType("boolean");
      sap[1].setValue("true");
      sap[1].setDescription("algorithm automatically builds the neural network");
      sap[1].setStatus(0);

      sap[2].setName("learningRate");
      sap[2].setType("double");
      sap[2].setValue("0.5");
      sap[2].setDescription("learning rate of backpropagation method");
      sap[2].setStatus(0);

      sap[3].setName("momentum");
      sap[3].setType("double");
      sap[3].setValue("0.3");
      sap[3].setDescription("momentum of backpropagation method, if used");
      sap[3].setStatus(0);

      sap[4].setName("maxNumberOfIterations");
      sap[4].setType("int");
      sap[4].setValue("400");
      sap[4].setDescription("Maximum number of iterations");
      sap[4].setStatus(0);

      sap[5].setName("maxError");
      sap[5].setType("double");
      sap[5].setValue("1.0E-35");
      sap[5].setDescription("maximum acceptable error of training the network");
      sap[5].setStatus(0);

      sapVec.addElement(sap);
    }
    if ( settings instanceof TimeSeriesMiningSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 3;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sap[0].setName("embeddingDimension");
      sap[0].setType("int");
      sap[0].setValue("5");
      sap[0].setDescription("embedding dimension");
      sap[0].setStatus(0);

      sap[1].setName("stepSize");
      sap[1].setType("int");
      sap[1].setValue("1");
      sap[1].setDescription("step size in time direction");
      sap[1].setStatus(0);

      sap[2].setName("singleApproximator");
      sap[2].setType("boolean");
      sap[2].setValue("true");
      sap[2].setDescription("use single or multiple approximators");
      sap[2].setStatus(0);

      sapVec.addElement(sap);
    }
    if ( settings instanceof StatisticsSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 4;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sap[0].setName("grouping");
      sap[0].setType("java.util.Vector");
      sap[0].setDescription("definition of grouping attributes");
      sap[0].setStatus(0);

      sap[1].setName("univariateTargetName");
      sap[1].setType("java.lang.String");
      sap[1].setDescription("name of univariate target attribute");
      sap[1].setStatus(1);

      sap[2].setName("multivariateTarget1Name");
      sap[2].setType("java.lang.String");
      sap[2].setDescription("name of first multivariate target attribute");
      sap[2].setStatus(0);

      sap[3].setName("multivariateTarget2Name");
      sap[3].setType("java.lang.String");
      sap[3].setDescription("name of second multivariate target attribute");
      sap[3].setStatus(0);

      sapVec.addElement(sap);
    }
    if ( settings instanceof ClusteringSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 3;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sap[0].setName("maxNumberOfClusters");
      sap[0].setType("int");
      sap[0].setValue("0");
      sap[0].setDescription("maximum number of clusters");
      sap[0].setStatus(0);

      sap[1].setName("clusterIdAttributeName");
      sap[1].setType("java.lang.String");
      sap[1].setValue("");
      sap[1].setDescription("name of ID attribute for output");
      sap[1].setStatus(1);

      sap[2].setName("distance");
      sap[2].setType("com.prudsys.pdm.Models.Clustering.Distance");
      sap[2].setDescription("object defining the metric");
      sap[2].setStatus(1);
      sap[2].setID("dist");

      sapVec.addElement(sap);

      /*------------------------------------------------------------------*/
      // Add distance object:
      boolean hierar = false;
      if ( ((ClusteringSettings) settings).getDistance() instanceof
          com.prudsys.pdm.Models.Clustering.Hierarchical.ClusterDistance)
        hierar = true;
      npar = 11;
      if (hierar) npar = npar + 1;
      String cstring = "";
      ServiceAlgorithmParameter[] csap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        csap[i] = new ServiceAlgorithmParameter();
        csap[i].setValue("");
        csap[i].setDomain(0);
        String ID = "dist" + String.valueOf(i);
        csap[i].setID( ID );
        csap[i].setChildIDs("");
        cstring = cstring + ID + " ";
      };
      sap[2].setChildIDs(cstring);

      csap[0].setName("type");
      csap[0].setType("int");
      csap[0].setValue("1");
      csap[0].setDescription("distance type (Euclidean, Chebychev, ...)");
      csap[0].setStatus(0);

      csap[1].setName("measureType");
      csap[1].setType("int");
      csap[1].setValue("10001");
      csap[1].setDescription("measure type (distance, similarity)");
      csap[1].setStatus(0);

      csap[2].setName("compareFunction");
      csap[2].setType("int");
      csap[2].setValue("101");
      csap[2].setDescription("compare function between attribute values");
      csap[2].setStatus(0);

      csap[3].setName("normalized");
      csap[3].setType("boolean");
      csap[3].setValue("false");
      csap[3].setDescription("use [0,1] normalization for all attributes");
      csap[3].setStatus(0);

      csap[4].setName("simMeasNormConst");
      csap[4].setType("double");
      csap[4].setValue("1.0");
      csap[4].setDescription("norming constant in distance invertation");
      csap[4].setStatus(0);

      csap[5].setName("minkPar");
      csap[5].setType("double");
      csap[5].setValue("2.0");
      csap[5].setDescription("value of Minkowski parameter");
      csap[5].setStatus(0);

      csap[6].setName("minAtt");
      csap[6].setType("double[]");
      csap[6].setValue("null");
      csap[6].setDescription("array of minimum values (required for normalization)");
      csap[6].setStatus(0);

      csap[7].setName("maxAtt");
      csap[7].setType("double[]");
      csap[7].setValue("null");
      csap[7].setDescription("array of maximum values (required for normalization)");
      csap[7].setStatus(0);

      csap[8].setName("fieldWeights");
      csap[8].setType("double[]");
      csap[8].setValue("null");
      csap[8].setDescription("array of attribute weights");
      csap[8].setStatus(0);

      csap[9].setName("minCompareFunction");
      csap[9].setType("double");
      csap[9].setValue("0");
      csap[9].setDescription("minimum compare function (PMML)");
      csap[9].setStatus(0);

      csap[10].setName("maxCompareFunction");
      csap[10].setType("double");
      csap[10].setValue("0");
      csap[10].setDescription("maximum compare function (PMML)");
      csap[10].setStatus(0);

      if (hierar) {
        csap[11].setName("clustDistType");
        csap[11].setType("int");
        csap[11].setValue("5");
        csap[11].setDescription("type of cluster distance (centroid, Ward, ...)");
        csap[11].setStatus(0);
      };

      sapVec.addElement(csap);
      /*------------------------------------------------------------------*/

    }
    if ( settings instanceof CDBasedClusteringSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 0;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sapVec.addElement(sap);
    }
    if ( settings instanceof HierarchicalClusteringSettings )
    {
      ServiceAlgorithmParameter[] sap = null;
      int npar = 0;
      sap = new ServiceAlgorithmParameter[npar];
      for (int i = 0; i < npar; i++) {
        sap[i] = new ServiceAlgorithmParameter();
        sap[i].setValue("");
        sap[i].setDomain(0);
        sap[i].setID("");
        sap[i].setChildIDs("");
      };

      sapVec.addElement(sap);