protected virtual double CalculateWeightUpdate(
Matrix <double> matrixX,
ILinearRegressionParams regressionParams,
double diffSums,
double currentWeightValue)
{
return(currentWeightValue - (((regressionParams.LearningRate * 1) / matrixX.RowCount) * diffSums));
}
public ILinearRegressionModel BuildModel(Matrix<double> matrixX, Vector<double> vectorY, ILinearRegressionParams linearRegressionParams)
{
var normalizationMatrix = Matrix<double>.Build.DenseIdentity(matrixX.ColumnCount, matrixX.ColumnCount) * regularizationConst;
var xT = matrixX.Transpose();
var xTx = xT.Multiply(matrixX) + normalizationMatrix;
var inversedXTX = xTx.Inverse();
var xTy = xT.Multiply(vectorY);
return new LinearRegressionModel(inversedXTX.Multiply(xTy));
}
protected override double CalculateWeightUpdate(
Matrix<double> matrixX,
ILinearRegressionParams regressionParams,
double diffSums,
double currentWeightValue)
{
var update = base.CalculateWeightUpdate(matrixX, regressionParams, diffSums, currentWeightValue);
var regularizationTerm = (regularizationConst / matrixX.RowCount) * currentWeightValue;
return update - regularizationTerm;
}
protected override double CalculateWeightUpdate(
Matrix <double> matrixX,
ILinearRegressionParams regressionParams,
double diffSums,
double currentWeightValue)
{
var update = base.CalculateWeightUpdate(matrixX, regressionParams, diffSums, currentWeightValue);
var regularizationTerm = (regularizationConst / matrixX.RowCount) * currentWeightValue;
return(update - regularizationTerm);
}
public ILinearRegressionModel BuildModel(Matrix<double> matrixX, Vector<double> vectorY, ILinearRegressionParams regressionParams)
{
var weightsVector = Vector<double>.Build.DenseOfArray(
Enumerable.Range(0, matrixX.ColumnCount)
.Select(fIdx => randomizer.NextDoubleInRange(weightsMin, weightsMax))
.ToArray());
double previousIterError = double.PositiveInfinity;
for (int iterNo = 0; iterNo < iterations; iterNo++)
{
var vectorYHat = matrixX.Multiply(weightsVector);
var diffVector = vectorYHat.Subtract(vectorY);
var currentIterError = diffVector.Select(Math.Abs).Sum();
if ((currentIterError <= stopLerningError) ||
currentIterError.AlmostEqual(previousIterError, 5))
{
break;
}
for (int weightIdx = 0; weightIdx < weightsVector.Count; weightIdx++)
{
var diffSums = 0.0;
for (int rowIdx = 0; rowIdx < matrixX.RowCount; rowIdx++)
{
diffSums += diffVector[rowIdx] * matrixX.Row(rowIdx)[weightIdx];
}
var newWeightValue = CalculateWeightUpdate(
matrixX,
regressionParams,
diffSums,
weightsVector[weightIdx]);
weightsVector[weightIdx] = newWeightValue;
}
previousIterError = currentIterError;
}
return new LinearRegressionModel(weightsVector);
}
public RegressionAndModelDecisionTreeLeafBuilder(ILinearRegressionModelBuilder modelBuilder, double learningRate = 0.05)
{
regressionModelBuilder = modelBuilder;
regressionParams = new LinearRegressionParams(learningRate);
}
public ILinearRegressionModel BuildModel(Matrix <double> matrixX, Vector <double> vectorY, ILinearRegressionParams regressionParams)
{
var weightsVector = Vector <double> .Build.DenseOfArray(
Enumerable.Range(0, matrixX.ColumnCount)
.Select(fIdx => randomizer.NextDoubleInRange(weightsMin, weightsMax))
.ToArray());
double previousIterError = double.PositiveInfinity;
for (int iterNo = 0; iterNo < iterations; iterNo++)
{
var vectorYHat = matrixX.Multiply(weightsVector);
var diffVector = vectorYHat.Subtract(vectorY);
var currentIterError = diffVector.Select(Math.Abs).Sum();
if ((currentIterError <= stopLerningError) ||
currentIterError.AlmostEqual(previousIterError, 5))
{
break;
}
for (int weightIdx = 0; weightIdx < weightsVector.Count; weightIdx++)
{
var diffSums = 0.0;
for (int rowIdx = 0; rowIdx < matrixX.RowCount; rowIdx++)
{
diffSums += diffVector[rowIdx] * matrixX.Row(rowIdx)[weightIdx];
}
var newWeightValue = CalculateWeightUpdate(
matrixX,
regressionParams,
diffSums,
weightsVector[weightIdx]);
weightsVector[weightIdx] = newWeightValue;
}
previousIterError = currentIterError;
}
return(new LinearRegressionModel(weightsVector));
}
public RegressionAndModelDecisionTreeLeafBuilder(ILinearRegressionModelBuilder modelBuilder, double learningRate = 0.05)
{
regressionModelBuilder = modelBuilder;
regressionParams = new LinearRegressionParams(learningRate);
}
protected virtual double CalculateWeightUpdate(
Matrix<double> matrixX,
ILinearRegressionParams regressionParams,
double diffSums,
double currentWeightValue)
{
return currentWeightValue - (((regressionParams.LearningRate * 1) / matrixX.RowCount) * diffSums);
}
public ILinearRegressionModel BuildModel(Matrix <double> matrixX, Vector <double> vectorY, ILinearRegressionParams linearRegressionParams)
{
var normalizationMatrix = Matrix <double> .Build.DenseIdentity(matrixX.ColumnCount, matrixX.ColumnCount) * regularizationConst;
var xT = matrixX.Transpose();
var xTx = xT.Multiply(matrixX) + normalizationMatrix;
var inversedXTX = xTx.Inverse();
var xTy = xT.Multiply(vectorY);
return(new LinearRegressionModel(inversedXTX.Multiply(xTy)));
}
