de.jungblut.math.minimize

Class ParticleSwarmOptimization

java.lang.Object

de.jungblut.math.minimize.AbstractMinimizer

de.jungblut.math.minimize.ParticleSwarmOptimization

All Implemented Interfaces:

Minimizer

public final class ParticleSwarmOptimization
extends AbstractMinimizer

Particle Swarm Optimization algorithm to minimize costfunctions. This works quite like GradientDescent, but for this to work we don't need to have a derivative, it is enough to provide the cost at a certain parameter position (theta). For additional information you can browse the wikipedia page: Particle swarm optimization article on Wikipedia

Author:

thomas.jungblut

Constructor Summary

Constructors

Constructor and Description
ParticleSwarmOptimization(int numParticles, double alpha, double beta, double phi, int numThreads)

Method Summary

Modifier and Type	Method and Description
de.jungblut.math.DoubleVector	minimize(CostFunction f, de.jungblut.math.DoubleVector pInput, int maxIterations, boolean verbose) Minimizes the given costfunction with the starting parameter theta.
static de.jungblut.math.DoubleVector	minimizeFunction(CostFunction f, de.jungblut.math.DoubleVector pInput, int numParticles, double alpha, double beta, double phi, int maxIterations, int numThreads, boolean verbose) Minimize a function using particle swarm optimization.

Methods inherited from class de.jungblut.math.minimize.AbstractMinimizer

addIterationCompletionCallback, onIterationFinished

Methods inherited from class java.lang.Object

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

Constructor Detail

ParticleSwarmOptimization

public ParticleSwarmOptimization(int numParticles,
                                 double alpha,
                                 double beta,
                                 double phi,
                                 int numThreads)

Method Detail

minimize

public final de.jungblut.math.DoubleVector minimize(CostFunction f,
                                                    de.jungblut.math.DoubleVector pInput,
                                                    int maxIterations,
                                                    boolean verbose)

Description copied from interface: Minimizer

Minimizes the given costfunction with the starting parameter theta.

Parameters:

f - the costfunction to minimize.

pInput - the starting parameters.

maxIterations - the number of iterations to do.

verbose - if TRUE it will print progress.

Returns:

the optimized theta parameters.

minimizeFunction

public static de.jungblut.math.DoubleVector minimizeFunction(CostFunction f,
                                                             de.jungblut.math.DoubleVector pInput,
                                                             int numParticles,
                                                             double alpha,
                                                             double beta,
                                                             double phi,
                                                             int maxIterations,
                                                             int numThreads,
                                                             boolean verbose)

Minimize a function using particle swarm optimization.

Parameters:

f - the cost function to minimize. Note that the returned gradient will be ignored, because it is not needed in this algorithm.

pInput - the initial starting point of the algorithm / particles. This is very important to choose, since this is considered a fast algorithm you could try out multiple starting points.

numParticles - how many particles to use.

alpha - personal memory weighting.

beta - group memory weighting.

phi - own velocity weighting (inertia).

maxIterations - how many iterations this algorithm should perform.

verbose - if true prints progress to STDOUT.

Returns:

a optimized parameter set for the costfunction.