/// <summary>Generate Linear Regression model based on a set of examples.</summary>
/// <param name="x">The Matrix to process.</param>
/// <param name="y">The Vector to process.</param>
/// <returns>Model.</returns>
public override IModel Generate(Matrix x, Vector y)
{
// create initial theta
Vector theta = Vector.Ones(x.Cols + 1);
Matrix copy = x.Copy();
// normalise features
for (int i = 0; i < copy.Cols; i++)
{
var j = FeatureNormalizer.FeatureScale(copy[i, VectorType.Col]);
for (int k = 0; k < copy.Rows; k++)
{
copy[k, i] = j[k];
}
}
// add intercept term
copy = copy.Insert(Vector.Ones(copy.Rows), 0, VectorType.Col);
// run gradient descent
var run = GradientDescent.Run(theta, copy, y, this.MaxIterations, this.LearningRate, new LinearCostFunction(), this.Lambda, new Regularization());
// once converged create model and apply theta
LinearRegressionModel model = new LinearRegressionModel(x.Mean(VectorType.Row), x.StdDev(VectorType.Row))
{
Descriptor = this.Descriptor,
Theta = run.Item2
};
return(model);
}
/// <summary>
/// Override to perform custom pre-processing steps on the raw Matrix data.
/// </summary>
/// <param name="X1">Matrix of initial states.</param>
/// <param name="y">Vector of action labels.</param>
/// <param name="X2">Matrix of transition states.</param>
/// <param name="r">Vector of reward values.</param>
/// <returns></returns>
public virtual void Preprocess(Matrix X1, Vector y, Matrix X2, Vector r)
{
FeatureProperties = Summary.Summarize(X1);
if (NormalizeFeatures)
{
if (FeatureNormalizer != null)
{
for (var i = 0; i < X1.Rows; i++)
{
var v1 = FeatureNormalizer.Normalize(X1[i, VectorType.Row], FeatureProperties);
X1[i, VectorType.Row] = v1;
if (X2 != null)
{
var v2 = FeatureNormalizer.Normalize(X2[i, VectorType.Row], FeatureProperties);
X2[i, VectorType.Row] = v2;
}
}
}
}
if (FeatureDiscretizer == null)
{
var temp = Matrix.VStack(X1, X2);
var bins = new double[temp.Cols];
for (var x = 0; x < X1.Cols; x++)
{
bins[x] = temp[x, VectorType.Col].Distinct().Count();
}
FeatureDiscretizer = new BinningDiscretizer(bins.ToVector());
FeatureDiscretizer.Initialize(X1, FeatureProperties);
}
}
/// <summary>
/// Override to perform custom pre-processing steps on the raw Matrix data.
/// </summary>
/// <param name="X">Matrix of examples.</param>
/// <returns></returns>
public virtual void Preprocess(Matrix X)
{
FeatureProperties = new Summary
{
Average = X.Mean(VectorType.Row),
StandardDeviation = X.StdDev(VectorType.Row),
Minimum = X.Min(VectorType.Row),
Maximum = X.Max(VectorType.Row),
Median = X.Median(VectorType.Row)
};
if (NormalizeFeatures)
{
if (FeatureNormalizer != null)
{
for (var i = 0; i < X.Rows; i++)
{
var vectors = FeatureNormalizer.Normalize(X[i, VectorType.Row], FeatureProperties);
for (var j = 0; j < X.Cols; j++)
{
X[i, j] = vectors[j];
}
}
}
}
}
/// <summary>
/// TODO The normalise.
/// </summary>
/// <param name="y">
/// TODO The y.
/// </param>
/// <returns>
/// </returns>
private Vector Normalise(Vector y)
{
for (var i = 0; i < y.Length; i++)
{
y[i] = FeatureNormalizer.FeatureScale(y[i], this.FeatureAverages[i], this.FeatureStandardDeviations[i]);
}
return(y.Insert(0, 1.0d));
}
public List <double> CompareToTemplate(List <SignatureFeatures> aFeaturesToCompare)
{
List <double> lScoreList = new List <double>();
foreach (var featureSet in aFeaturesToCompare)
{
lScoreList.Add(CalculateScores(mTemplate, FeatureNormalizer.NormalizeFeaturesSingleSignature(featureSet, mNormalized.NormalizedWith)));
}
return(lScoreList);
}
/// <summary>
/// Preprocessed the input vector.
/// </summary>
/// <param name="x">Input vector.</param>
/// <returns>Vector.</returns>
protected void Preprocess(Vector x)
{
if (NormalizeFeatures)
{
var xp = FeatureNormalizer.Normalize(x, FeatureProperties);
for (var i = 0; i < x.Length; i++)
{
x[i] = xp[i];
}
}
}
public Detector(List <SignatureFeatures> aFeatures)
{
mNormalized = FeatureNormalizer.NormalizeFeatures(aFeatures);
foreach (var featureSet in mNormalized.ListOfFeatureSets)
{
foreach (var feature in featureSet)
{
mTemplate[feature.Key] += feature.Value;
}
}
foreach (var feature in mTemplate.ToList())
{
mTemplate[feature.Key] /= mNormalized.ListOfFeatureSets.Count;
}
}
