T
- Data type of input spacepublic abstract class GaussianKernel<T> extends Kernel<T>
Base class for Gaussian Kernels in the form of k(x1, x2) = exp(-gamme * dist(x1, x2))
The distance used if defined in specific subclasses.
Constructor and Description |
---|
GaussianKernel() |
Modifier and Type | Method and Description |
---|---|
abstract double |
distanceValueOf(T t1,
T t2)
Tells the inner distance between two samples used by this Gaussian kernel.
|
double[][] |
getDistanceMatrix(java.util.List<TrainingSample<T>> l)
Tells the distance matrix for a specified list of samples.
This is a threaded operation. |
abstract double |
getGamma()
Tells exponential coefficient
|
abstract void |
setGamma(double gamma)
Sets exponential coefficient.
|
getKernelMatrix, getNormalizedKernelMatrix, normalizedValueOf, setName, toString, valueOf, valueOf
public abstract void setGamma(double gamma)
gamma
- gamma coefficientpublic abstract double getGamma()
public abstract double distanceValueOf(T t1, T t2)
t1
- first samplet2
- second samplepublic double[][] getDistanceMatrix(java.util.List<TrainingSample<T>> l)
l
- the list of samples