jsat.distributions.empirical.kernelfunc

Class BiweightKF

java.lang.Object

jsat.distributions.empirical.kernelfunc.BiweightKF

All Implemented Interfaces:

Serializable, KernelFunction

public class BiweightKF
extends Object
implements KernelFunction

Author:

Edward Raff

See Also:

Serialized Form

Method Summary

Modifier and Type	Method and Description
double	cutOff() As the value of |u| for the kernel function approaches infinity, the value of k(u) approaches zero.
static BiweightKF	getInstance() Returns the singleton instance of this class
double	intK(double u) Computes the value of the finite integral from -Infinity up to the value u, of the function given by KernelFunction.k(double)
double	k(double u) Returns the weight to be applied to a sample for the normalized distance of two data points.
double	k2() Returns the variance of the kernel function
double	kPrime(double u) Returns the value of the derivative at a point, k'(u)
String	toString() Returns the name of this kernel function

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Method Detail

getInstance

public static BiweightKF getInstance()

Returns the singleton instance of this class

Returns:

the instance of this class

k

public double k(double u)

Description copied from interface: KernelFunction

Returns the weight to be applied to a sample for the normalized distance of two data points.

Specified by:

k in interface KernelFunction

Parameters:

u - the distance of the data points

Returns:

the value in [0, 1) of the amount of weight to give to the sample based on its distance

intK

public double intK(double u)

Description copied from interface: KernelFunction

Computes the value of the finite integral from -Infinity up to the value u, of the function given by KernelFunction.k(double)

Specified by:

intK in interface KernelFunction

Parameters:

u - the distance of the data points

Returns:

the value of the integration

k2

public double k2()

Description copied from interface: KernelFunction

Returns the variance of the kernel function

Specified by:

k2 in interface KernelFunction

Returns:

the variance of the kernel function

cutOff

public double cutOff()

Description copied from interface: KernelFunction

As the value of |u| for the kernel function approaches infinity, the value of k(u) approaches zero. This function returns the minimal absolute value of u for which k(u) returns 0

Specified by:

cutOff in interface KernelFunction

Returns:

the first value for which k(u) = 0

kPrime

public double kPrime(double u)

Description copied from interface: KernelFunction

Returns the value of the derivative at a point, k'(u)

Specified by:

kPrime in interface KernelFunction

Parameters:

u - the distance of the data points

Returns:

the value of the derivative at u

toString

public String toString()

Description copied from interface: KernelFunction

Returns the name of this kernel function

Specified by:

toString in interface KernelFunction

Overrides:

toString in class Object

Returns:

the name of this kernel function