Merge pull request #649 from puffin444/master

ARD and Gradient Search Vector Learning
shogun-toolbox · Jul 18, 2012 · 1e888a0 · 1e888a0
2 parents 7565032 + bcb7e43
commit 1e888a0
Show file tree

Hide file tree

Showing 19 changed files with 984 additions and 466 deletions.
diff --git a/examples/undocumented/libshogun/Makefile b/examples/undocumented/libshogun/Makefile
@@ -38,6 +38,7 @@ TARGETS = basic_minimal \
 		  regression_gaussian_process_gaussian \
 		  regression_gaussian_process_sum \
 		  regression_gaussian_process_product \
+		  regression_gaussian_process_ard \
 		  modelselection_model_selection_parameters_test \
 		  modelselection_parameter_tree \
 		  modelselection_apply_parameter_tree \

diff --git a/examples/undocumented/libshogun/regression_gaussian_process_ard.cpp b/examples/undocumented/libshogun/regression_gaussian_process_ard.cpp
@@ -0,0 +1,224 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * Written (W) 2012 Jacob Walker
+ */
+
+#include <shogun/base/init.h>
+#include <shogun/labels/RegressionLabels.h>
+#include <shogun/features/DenseFeatures.h>
+#include <shogun/kernel/LinearARDKernel.h>
+#include <shogun/mathematics/Math.h>
+#include <shogun/regression/gp/ExactInferenceMethod.h>
+#include <shogun/regression/gp/GaussianLikelihood.h>
+#include <shogun/regression/gp/ZeroMean.h>
+#include <shogun/regression/GaussianProcessRegression.h>
+#include <shogun/evaluation/GradientEvaluation.h>
+#include <shogun/modelselection/GradientModelSelection.h>
+#include <shogun/modelselection/ModelSelectionParameters.h>
+#include <shogun/modelselection/ParameterCombination.h>
+#include <shogun/evaluation/GradientCriterion.h>
+
+using namespace shogun;
+
+int32_t num_vectors=4;
+int32_t dim_vectors=3;
+
+void build_matrices(SGMatrix<float64_t>& test, SGMatrix<float64_t>& train,
+		    CRegressionLabels* labels)
+{
+ 	/*Fill Matrices with random nonsense*/
+	train[0] = -1;
+	train[1] = -1;
+	train[2] = -1;
+	train[3] = 1;
+	train[4] = 1;
+	train[5] = 1;
+	train[6] = -10;
+	train[7] = -10;
+	train[8] = -10;
+	train[9] = 3;
+	train[10] = 2;
+	train[11] = 1;
+
+	for (int32_t i=0; i<num_vectors*dim_vectors; i++)
+	    test[i]=i*sin(i)*.96; 
+
+	/* create labels, two classes */
+	for (index_t i=0; i<num_vectors; ++i)
+	{
+		if(i%2 == 0) labels->set_label(i, 1);
+		else labels->set_label(i, -1);
+	}
+}
+
+CModelSelectionParameters* build_tree(CInferenceMethod* inf, 
+				      CLikelihoodModel* lik, CKernel* kernel,
+				      SGVector<float64_t>& weights)
+{
+  	CModelSelectionParameters* root=new CModelSelectionParameters();
+
+	CModelSelectionParameters* c1 =
+			new CModelSelectionParameters("inference_method", inf);
+	root->append_child(c1);
+
+	CModelSelectionParameters* c2 =
+			new CModelSelectionParameters("likelihood_model", lik);
+	c1->append_child(c2);
+
+	CModelSelectionParameters* c3=new CModelSelectionParameters("sigma");
+	c2->append_child(c3);
+	c3->build_values(1.0, 4.0, R_LINEAR);
+
+        CModelSelectionParameters* c4=new CModelSelectionParameters("scale");
+        c1->append_child(c4);
+        c4->build_values(1.0, 1.0, R_LINEAR);
+
+	CModelSelectionParameters* c5 =
+			new CModelSelectionParameters("kernel", kernel);
+	c1->append_child(c5);
+
+	CModelSelectionParameters* c6 =
+			new CModelSelectionParameters("weights");
+	c5->append_child(c6);
+	c6->build_values_sgvector(0.001, 4.0, R_LINEAR, &weights);
+
+	return root;
+}
+
+int main(int argc, char **argv)
+{
+	init_shogun_with_defaults();
+
+	/* create some data and labels */
+	SGMatrix<float64_t> matrix =
+			SGMatrix<float64_t>(dim_vectors, num_vectors);
+
+	SGVector<float64_t> weights(dim_vectors);
+
+	SGMatrix<float64_t> matrix2 =
+			SGMatrix<float64_t>(dim_vectors, num_vectors);
+
+	CRegressionLabels* labels=new CRegressionLabels(num_vectors);
+
+	build_matrices(matrix2, matrix, labels);
+
+	/* create training features */
+	CDenseFeatures<float64_t>* features=new CDenseFeatures<float64_t> ();
+	features->set_feature_matrix(matrix);
+
+	/* create testing features */
+	CDenseFeatures<float64_t>* features2=new CDenseFeatures<float64_t> ();
+	features2->set_feature_matrix(matrix2);
+
+	SG_REF(features);
+	SG_REF(features2);
+
+	SG_REF(labels);
+
+	/*Allocate our Kernel*/
+	CLinearARDKernel* test_kernel = new CLinearARDKernel(10);
+
+	test_kernel->init(features, features);
+
+	/*Allocate our mean function*/
+	CZeroMean* mean = new CZeroMean();
+
+	/*Allocate our likelihood function*/
+	CGaussianLikelihood* lik = new CGaussianLikelihood();
+
+	/*Allocate our inference method*/
+	CExactInferenceMethod* inf =
+			new CExactInferenceMethod(test_kernel, 
+						  features, mean, labels, lik);
+
+	SG_REF(inf);
+
+	/*Finally use these to allocate the Gaussian Process Object*/
+	CGaussianProcessRegression* gp =
+			new CGaussianProcessRegression(inf, features, labels);
+
+	SG_REF(gp);
+
+	/*Build the parameter tree for model selection*/
+	CModelSelectionParameters* root = build_tree(inf, lik, test_kernel, 
+						     weights);
+
+	/*Criterion for gradient search*/
+	CGradientCriterion* crit = new CGradientCriterion();
+
+	/*This will evaluate our inference method for its derivatives*/
+	CGradientEvaluation* grad=new CGradientEvaluation(gp, features, labels,
+			crit);
+
+	grad->set_function(inf);
+
+	gp->print_modsel_params();
+
+	root->print_tree();
+
+	/* handles all of the above structures in memory */
+	CGradientModelSelection* grad_search=new CGradientModelSelection(
+			root, grad);
+
+	/* set autolocking to false to get rid of warnings */
+	grad->set_autolock(false);
+
+	/*Search for best parameters*/
+	CParameterCombination* best_combination=grad_search->select_model(true);
+
+	/*Output all the results and information*/
+	if (best_combination)
+	{
+		SG_SPRINT("best parameter(s):\n");
+		best_combination->print_tree();
+
+		best_combination->apply_to_machine(gp);
+	}
+
+	CGradientResult* result=(CGradientResult*)grad->evaluate();
+
+	if(result->get_result_type() != GRADIENTEVALUATION_RESULT)
+		SG_SERROR("Evaluation result not a GradientEvaluationResult!");
+
+	result->print_result();
+
+	SGVector<float64_t> alpha = inf->get_alpha();
+	SGVector<float64_t> labe = labels->get_labels();
+	SGVector<float64_t> diagonal = inf->get_diagonal_vector();
+	SGMatrix<float64_t> cholesky = inf->get_cholesky();
+	gp->set_return_type(CGaussianProcessRegression::GP_RETURN_COV);
+
+	CRegressionLabels* covariance = gp->apply_regression(features);
+
+	gp->set_return_type(CGaussianProcessRegression::GP_RETURN_MEANS);
+	CRegressionLabels* predictions = gp->apply_regression();
+
+	alpha.display_vector("Alpha Vector");
+	labe.display_vector("Labels");
+	diagonal.display_vector("sW Matrix");
+	covariance->get_labels().display_vector("Predicted Variances");
+	predictions->get_labels().display_vector("Mean Predictions");
+	cholesky.display_matrix("Cholesky Matrix L");
+	matrix.display_matrix("Training Features");
+	matrix2.display_matrix("Testing Features");
+
+	/*free memory*/
+	SG_UNREF(features);
+	SG_UNREF(features2);
+	SG_UNREF(predictions);
+	SG_UNREF(covariance);
+	SG_UNREF(labels);
+	SG_UNREF(inf);
+	SG_UNREF(gp);
+	SG_UNREF(grad_search);
+	SG_UNREF(best_combination);
+	SG_UNREF(result);
+
+	exit_shogun();
+
+	return 0;
+}