Added RandomForests

config/Algorithms.xml

@@ -505,7 +505,11 @@
 		</algorithm>
 
 		<algorithm name="SMO (no complexity estimation)" class="ecst.algorithm.classification.SMOAdapter">
-
+
+		</algorithm>
+
+		<algorithm name="Random Forest (no complexity estimation)" class="ecst.algorithm.classification.RandomForestAdapter">
+
 		</algorithm>
 
 		<algorithm name="Naive Bayes" class="ecst.algorithm.classification.NaiveBayesAdapter">

src/ecst/algorithm/classification/RandomForestAdapter.java

@@ -0,0 +1,128 @@
+package ecst.algorithm.classification;
+
+import ecst.algorithm.ClassificationAlgorithm;
+import ecst.algorithm.analysis.DynamicMultiplier;
+import ecst.algorithm.parameter.Parameter;
+import ecst.combiner.PipelineData;
+import weka.classifiers.Classifier;
+import weka.classifiers.trees.RandomForest;
+
+public class RandomForestAdapter extends ClassificationAlgorithm {
+
+	private Parameter seed;
+	private Parameter maxDepth;
+	private Parameter iterations;
+	private Parameter randomAttributes;
+
+	// WOULD REQUIRE WEKA 3.8
+	// private Parameter bagSize;
+	// private Parameter calculateOutOfBagError;
+	// private Parameter storeOutOfBagPredictions;
+	// private Parameter outputOutOfBagStatistics;
+	// private Parameter printClassifiers;
+	// private Parameter attributeImportance;
+	// private Parameter slots;
+	// private Parameter instancesPerLeaf;
+	// private Parameter minimumVariance;
+	// private Parameter foldsForBackfitting;
+	// private Parameter unclassifiedInstances;
+	// private Parameter breakTieRandom;
+	// private Parameter debug;
+	// private Parameter checkCapabilities;
+	// private Parameter decimalPlaces;
+	// private Parameter batchSize;
+
+	@Override
+	protected void initClassifierParameters() {
+		iterations = new Parameter(100, "Number of iterations", Parameter.TYPE.INTEGER, "-I");
+		randomAttributes = new Parameter(0, "Number of attributes to investigate randomly", Parameter.TYPE.INTEGER,
+				"-K");
+		seed = new Parameter(1, "Seed for random number generator", Parameter.TYPE.INTEGER, "-S");
+		maxDepth = new Parameter(0, "Maximum depth of the tree", Parameter.TYPE.INTEGER, "-depth");
+
+		// WOULD REQUIRE WEKA 3.8
+		// bagSize = new Parameter(100, "Bag size in percent", Parameter.TYPE.INTEGER,
+		// "-P");
+		// calculateOutOfBagError = new Parameter(false, "Calculate out-of-bag error",
+		// Parameter.TYPE.BOOLEAN, "-O");
+		// storeOutOfBagPredictions = new Parameter(false, "Store out-of-bag
+		// predictions", Parameter.TYPE.BOOLEAN,
+		// "-store-out-of-bag-predictions");
+		// outputOutOfBagStatistics = new Parameter(false, "Output out-of-bag complexity
+		// statistics",
+		// Parameter.TYPE.BOOLEAN, "-output-out-of-bag-complexity-statistics");
+		// printClassifiers = new Parameter(false, "Print individual classifiers",
+		// Parameter.TYPE.BOOLEAN, "-print");
+		// attributeImportance = new Parameter(false, "Compute and print attribute
+		// importance", Parameter.TYPE.BOOLEAN,
+		// "-attribute-importance");
+		// slots = new Parameter(1, "Number of execution slots", Parameter.TYPE.INTEGER,
+		// "-num-slots");
+		// instancesPerLeaf = new Parameter(1, "Minimum number of instances per leaf",
+		// Parameter.TYPE.INTEGER, "-M");
+		// minimumVariance = new Parameter(1.0e-3, "Minimum variance for split",
+		// Parameter.TYPE.DOUBLE, "-V");
+		// foldsForBackfitting = new Parameter(0, "Number of folds for backfitting",
+		// Parameter.TYPE.INTEGER, "-N");
+		// unclassifiedInstances = new Parameter(false, "Allow unclassified instances",
+		// Parameter.TYPE.BOOLEAN, "-U");
+		// breakTieRandom = new Parameter(false, "Break ties randomly",
+		// Parameter.TYPE.BOOLEAN, "-B");
+		// debug = new Parameter(false, "Debug mode", Parameter.TYPE.BOOLEAN,
+		// "-output-debug-info");
+		// checkCapabilities = new Parameter(false, "Do not check capabilities",
+		// Parameter.TYPE.BOOLEAN,
+		// "-do-not-check-capabilities");
+		// decimalPlaces = new Parameter(2, "Decimal places for the output of numbers",
+		// Parameter.TYPE.INTEGER,
+		// "-num-decimal-places");
+		// batchSize = new Parameter(100.0, "Batch size", Parameter.TYPE.DOUBLE,
+		// "-batch-size");
+	}
+
+	@Override
+	public Parameter[] getParameters() {
+		return new Parameter[] { iterations, randomAttributes, seed, maxDepth };
+
+		// WOULD REQUIRE WEKA 3.8
+		// return new Parameter[] { bagSize, calculateOutOfBagError,
+		// storeOutOfBagPredictions, outputOutOfBagStatistics,
+		// printClassifiers, attributeImportance, iterations, slots, randomAttributes,
+		// instancesPerLeaf,
+		// minimumVariance, seed, maxDepth, foldsForBackfitting, unclassifiedInstances,
+		// breakTieRandom, debug,
+		// checkCapabilities, decimalPlaces, batchSize };
+	}
+
+	@Override
+	public Class<? extends Object> getImplementingClass() {
+		return RandomForest.class;
+	}
+
+	@Override
+	protected Parameter[] getGridSearchParameters() {
+		return null;
+	}
+
+	@Override
+	protected void postprocess(Classifier classifier, PipelineData data) throws Exception {
+	}
+
+	@Override
+	protected void analyzeSystem(PipelineData data, Classifier classifier) throws Exception {
+	}
+
+	@Override
+	protected void addDependencies() {
+	}
+
+	@Override
+	protected void setMultiplier(DynamicMultiplier multiplier) {
+	}
+
+	@Override
+	public String modelToXML(PipelineData data) {
+		return null;
+	}
+
+}

src/ecst/view/ECST.java

@@ -44,7 +44,7 @@ public class ECST extends JFrame {
 	private static final long serialVersionUID = 1L;
 
 	public static final String PROGRAMM_NAME = "Embedded Classifcation Software Toolbox";
-	public static final String VERSION = "1.7";
+	public static final String VERSION = "1.7.1";
 	public static final String URL = "http://www.tinyurl.com/ecst-project";
 
 	private static final String CONFIGURATION_OPTION_SHORT = "c";