Using the JDM API - Java Data Mining: Strategy, Standard, and Practice - page 219

48. for(int i=0; i < (layerIds.length-1); i++) {
49. int[] parentNeuronIds = modelDetails.getNeuronIdentifiers(layerIds[i]);
50. int[] childNeuronIds = modelDetails.getNeuronIdentifiers(layerIds[i+1]);
51. int parentLayerId = layerIds[i];
52. for(int j=0; j < parentNeuronIds.length; j++) {
53. for(int k=0; k < childNeuronIds.length; k++) {
54. System.out.println("Wight for the parent layer " + parentLayerId +
55. ", parent neuron id " + parentNeuronIds[j] +
56. ", child neuron id " + childNeuronIds[k] + " is " +
57. modelDetails.getWeight(parentLayerId, parentNeuronIds[j], childNeuronIds[k]));
58. }}}}
59.}

Table 9-13 javax.datamining.modeldetails interfaces for classification function

javax.datamining.modeldetails package

Decision Tree: javax.datamining.modeldetails.tree package

TreeModel TreeModelDetail contains details of a tree model, such as, tree struc-

ture, nodes and rule associated with each node.

TreeNode A TreeNode characterizes a partition of a multidimensional dataset.

Naïve Bayes: javax.datamining.algorithm.naivebayes

NaiveBayesModelDetail NaiveBayesModelDetail contains the counts of the occurrences of

each target value, and counts of the co-occurrences of predictor
values with target values.

Support Vector Machine (SVM): javax.datamining.algorithm.svm.classification

SVMModelDetail SVMModelDetail is the superinterface for classification and

regression SVM model details.

SVMClassificationModelDetail An SVMModelDetail contains details of the SVM classification

model. SVM model details contain the support vector details, such as
number of bounded and unbounded vectors, coefficients associated
with the target attribute value and predictor attribute value pairs.

SVMRegressionModelDetail An SVMModelDetail contains details of the SVM regression model.

Similar to SVM classification regression model detail contains the
support vector details. The difference is SVM regression contains
coefficients associated with the predictor attributes and their values
as target is continuous number in case of regression.

Feed Forward NeuralNet: javax.datamining.algorithm.feedforwardneuralnet

NeuralNetworkModelDetail NeuralNetworkModelDetail captures the detailed representation of a

fully connected, multilayer, feed-forward neural network model.