/// <summary>Generate Logistic 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
var copy = x.Copy();
copy = FeatureDimensions.IncreaseDimensions(copy, this.PolynomialFeatures);
// add intercept term
copy = copy.Insert(Vector.Ones(copy.Rows), 0, VectorType.Col);
var theta = Vector.Ones(copy.Cols);
var run = GradientDescent.Run(
theta,
copy,
y,
this.MaxIterations,
this.LearningRate,
new LogisticCostFunction(),
this.Lambda,
new Regularization());
var model = new LogisticRegressionModel()
{
Descriptor = this.Descriptor, Theta = run.Item2, LogisticFunction = new Logistic(),
PolynomialFeatures = this.PolynomialFeatures
};
return(model);
}
/// <summary>
/// Create a prediction based on the learned Theta values and the supplied test item.
/// </summary>
/// <param name="y">Training record</param>
/// <returns></returns>
public override double Predict(Vector y)
{
var tempy = this.PolynomialFeatures > 0
? FeatureDimensions.IncreaseDimensions(y, this.PolynomialFeatures)
: y;
tempy = tempy.Insert(0, 1.0);
return(this.LogisticFunction.Compute((tempy * this.Theta).ToDouble()) >= 0.5 ? 1d : 0d);
}
public static FeatureDimensions dynamic_cast(FeatureBase arg0)
{
IntPtr cPtr = libtischPINVOKE.FeatureDimensions_dynamic_cast(FeatureBase.getCPtr(arg0));
FeatureDimensions ret = (cPtr == IntPtr.Zero) ? null : new FeatureDimensions(cPtr, false);
if (libtischPINVOKE.SWIGPendingException.Pending)
{
throw libtischPINVOKE.SWIGPendingException.Retrieve();
}
return(ret);
}
internal static HandleRef getCPtr(FeatureDimensions obj)
{
return((obj == null) ? new HandleRef(null, IntPtr.Zero) : obj.swigCPtr);
}
internal static HandleRef getCPtr(FeatureDimensions obj) {
return (obj == null) ? new HandleRef(null, IntPtr.Zero) : obj.swigCPtr;
}
