/// <summary>Train a multinomial classifier off of the provided dataset.</summary>
/// <param name="dataset">The dataset to train the classifier off of.</param>
/// <returns>A classifier.</returns>
public static IClassifier <string, string> TrainMultinomialClassifier(GeneralDataset <string, string> dataset, int featureThreshold, double sigma)
{
// Set up the dataset and factory
log.Info("Applying feature threshold (" + featureThreshold + ")...");
dataset.ApplyFeatureCountThreshold(featureThreshold);
log.Info("Randomizing dataset...");
dataset.Randomize(42l);
log.Info("Creating factory...");
LinearClassifierFactory <string, string> factory = InitFactory(sigma);
// Train the final classifier
log.Info("BEGIN training");
LinearClassifier <string, string> classifier = factory.TrainClassifier(dataset);
log.Info("END training");
// Debug
KBPRelationExtractor.Accuracy trainAccuracy = new KBPRelationExtractor.Accuracy();
foreach (IDatum <string, string> datum in dataset)
{
string guess = classifier.ClassOf(datum);
trainAccuracy.Predict(Java.Util.Collections.Singleton(guess), Java.Util.Collections.Singleton(datum.Label()));
}
log.Info("Training accuracy:");
log.Info(trainAccuracy.ToString());
log.Info(string.Empty);
// Return the classifier
return(classifier);
}
public virtual void TestAccuracyNoRelation()
{
KBPRelationExtractor.Accuracy accuracy = new KBPRelationExtractor.Accuracy();
accuracy.Predict(new HashSet <string>(Arrays.AsList("a")), new HashSet <string>(Arrays.AsList("a")));
accuracy.Predict(new HashSet <string>(Arrays.AsList("a")), new HashSet <string>(Arrays.AsList("no_relation")));
accuracy.Predict(new HashSet <string>(Arrays.AsList("no_relation")), new HashSet <string>(Arrays.AsList("b")));
accuracy.Predict(new HashSet <string>(Arrays.AsList("b")), new HashSet <string>(Arrays.AsList("no_relation")));
accuracy.Predict(new HashSet <string>(Arrays.AsList("b")), new HashSet <string>(Arrays.AsList("b")));
accuracy.Predict(new HashSet <string>(Arrays.AsList("b")), new HashSet <string>(Arrays.AsList("b")));
NUnit.Framework.Assert.AreEqual(accuracy.Precision("a"), 1e-10, 0.5);
NUnit.Framework.Assert.AreEqual(accuracy.Recall("a"), 1e-10, 1.0);
NUnit.Framework.Assert.AreEqual(accuracy.F1("a"), 1e-10, 2.0 * 1.0 * 0.5 / (1.0 + 0.5));
NUnit.Framework.Assert.AreEqual(accuracy.Precision("b"), 1e-10, 2.0 / 3.0);
NUnit.Framework.Assert.AreEqual(accuracy.Recall("b"), 1e-10, 2.0 / 3.0);
NUnit.Framework.Assert.AreEqual(accuracy.PrecisionMicro(), 1e-10, 3.0 / 5.0);
NUnit.Framework.Assert.AreEqual(accuracy.PrecisionMacro(), 1e-10, 7.0 / 12.0);
NUnit.Framework.Assert.AreEqual(accuracy.RecallMicro(), 1e-10, 3.0 / 4.0);
NUnit.Framework.Assert.AreEqual(accuracy.RecallMacro(), 1e-10, 5.0 / 6.0);
}