public Learn ( MathNet.Numerics.LinearAlgebra.Double.SparseMatrix input ) : void | ||
input | MathNet.Numerics.LinearAlgebra.Double.SparseMatrix | |
Résultat | void |
public void ErrorWhenLearningAfterInferenceMode() { var node = new SpatialNode2DGaussian(); var mat = new SparseMatrix(4, 4, 4.0); var learnAfterInferFails = false; var learnAfterTBInferFails = false; node.Learn(mat); node.Infer(mat); try { node.Learn(mat); } catch (HtmRuleException e) { learnAfterInferFails = true; Debug.WriteLine(e.Message); } node.TimeInfer(mat); try { node.Learn(mat); } catch (HtmRuleException e) { learnAfterTBInferFails = true; Debug.WriteLine(e.Message); } Assert.IsTrue(learnAfterInferFails); Assert.IsTrue(learnAfterTBInferFails); }
public void ErrorWhenLearningVaryingSizeInputs() { var node = new SpatialNode2DGaussian(); node.Learn(new SparseMatrix(5, 5, 3.0)); try { node.Learn(new SparseMatrix(4, 4, 2.0)); Assert.Inconclusive("Should have fired an exception"); } catch (Exception e) { Debug.WriteLine(e.Message); Assert.IsInstanceOfType(e, typeof(HtmRuleException)); } }
public void DontLearnBlankInputs() { var node = new SpatialNode2DGaussian(); node.Learn(new SparseMatrix(4)); Assert.AreEqual(0, node.CoincidencesFrequencies.Count); }
public void InferCorrectlyOnIdenticalMatricesWithMaxDistance0() { var node = new SpatialNode2DGaussian(); var ones = new SparseMatrix(4, 5, 1.0); var twos = new SparseMatrix(4, 5, 2.0); var expectedLower = Math.Exp(-(ones.RowCount * ones.ColumnCount) / (2 * node.SquaredSigma)) - 0.0001; var expectedUpper = Math.Exp(-(ones.RowCount * ones.ColumnCount) / (2 * node.SquaredSigma)) + 0.0001; node.Learn(ones); node.Learn(twos); var output = node.Infer(ones); Assert.IsTrue(output[0] == 1.0 && output[1] > expectedLower && output[1] < expectedUpper); }
public void TwoIdenticalInputsYieldOneCoincidenceOfFrequency2WithMaxDistance0() { // Arrange double maxDistance = 0.0; SpatialNode2DGaussian node = new SpatialNode2DGaussian(maxDistance, 0, 1); int nbInputs = 2; int width = 4; int height = 4; // Act for (int i = 0; i < nbInputs; ++i) node.Learn(new SparseMatrix(width, height, 1.0)); // Assert Assert.AreEqual(1, node.CoincidencesFrequencies.Count); Assert.AreEqual(nbInputs, node.CoincidencesFrequencies[new SparseMatrix(width, height, 1.0)]); }
public void StopLearningWhenMaxOutputSizeReached() { int max = 1; var node = new SpatialNode2DGaussian(0.0, 0.0, max); var ones = SparseMatrix.Identity(5); var twos = SparseMatrix.Identity(5) * 2.0; node.Learn(ones); node.Learn(twos); Assert.AreEqual(max, node.CoincidencesFrequencies.Count); }
public void Several5By5InputsYieldSeveralCoincidenceWithMaxDistance4_999() { var node = new SpatialNode2DGaussian(4.999); var ones = new SparseMatrix(5, 5, 1.0); var twos = new SparseMatrix(5, 5, 2.0); node.Learn(ones); node.Learn(twos); Assert.AreNotEqual(1, node.CoincidencesFrequencies.Count); Assert.AreEqual(2, node.CoincidencesFrequencies.Count); }
public void Several1DistantInputsYieldOneCoincidenceWithMaxDistance1() { // Arrange double maxDistance = 1.0; SpatialNode2DGaussian node = new SpatialNode2DGaussian(maxDistance); var matrices = new List<SparseMatrix>(); matrices.Add(new SparseMatrix(new double[,] { { 3.0, 2.0, 1.0, 0.0 } })); matrices.Add(new SparseMatrix(new double[,] { { 3.0, 2.0, 0.0, 0.0 } })); matrices.Add(new SparseMatrix(new double[,] { { 3.0, 2.0, 2.0, 0.0 } })); matrices.Add(new SparseMatrix(new double[,] { { 3.0, 1.0, 1.0, 0.0 } })); matrices.Add(new SparseMatrix(new double[,] { { 2.0, 2.0, 1.0, 0.0 } })); int nbInputs = matrices.Count ; // Act //node.Learn(); foreach (var matrix in matrices) node.Learn(matrix); // Assert Assert.AreEqual(1, node.CoincidencesFrequencies.Count); Assert.AreNotEqual(nbInputs, node.CoincidencesFrequencies.Count); }
public void RandomInputsYieldOnlyOneCoincidenceWithLargeMaxDistance() { // Arrange double maxDistance = 10000.0; SpatialNode2DGaussian node = new SpatialNode2DGaussian(maxDistance); int nbInputs = 100; int width = 4; int height = 4; var random = new Random(10); // Act for (int i = 0; i < nbInputs; ++i) node.Learn(new SparseMatrix(width, height, random.NextDouble())); // Assert Assert.AreEqual(1, node.CoincidencesFrequencies.Count); Assert.AreNotEqual(nbInputs, node.CoincidencesFrequencies.Count); }