org.aksw.limes.core.ml.algorithm.euclid

Class MeshBasedSelfConfigurator

java.lang.Object

org.aksw.limes.core.ml.algorithm.euclid.LinearSelfConfigurator

org.aksw.limes.core.ml.algorithm.euclid.BooleanSelfConfigurator

org.aksw.limes.core.ml.algorithm.euclid.MeshBasedSelfConfigurator

All Implemented Interfaces:

ISelfConfigurator

public class MeshBasedSelfConfigurator
extends BooleanSelfConfigurator

Author:

Kevin Dreßler, Mohamed Sherif (sherif@informatik.uni-leipzig.de), Klaus Lyko (lyko@informatik.uni-leipzig.de)

Nested Class Summary

Nested classes/interfaces inherited from class org.aksw.limes.core.ml.algorithm.euclid.LinearSelfConfigurator

LinearSelfConfigurator.QMeasureType, LinearSelfConfigurator.Strategy

Field Summary

Fields

Modifier and Type	Field and Description
static String	MAX_STRATEGY
static String	MIN_STRATEFY

Fields inherited from class org.aksw.limes.core.ml.algorithm.euclid.LinearSelfConfigurator

asked, ITERATIONS_MAX, kappa, learningRate, min_coverage, MIN_THRESHOLD, source, STRICT, target

Constructor Summary

Constructors

Constructor and Description
MeshBasedSelfConfigurator(ACache source, ACache target) Basic constructor uses default values for minCoverage and beta.
MeshBasedSelfConfigurator(ACache source, ACache target, double minCoverage) Constructor

Method Summary

Modifier and Type	Method and Description
static List<Double>	copy(List<Double> l)
static List<List<Double>>	generateCoordinates(List<Double> min, List<Double> max, int n) Generates a mesh over the space defined by the mins and max set.
ComplexClassifier	getHillTop(int n, List<SimpleClassifier> sc)
ComplexClassifier	getHillTop(List<Double> min, List<Double> max, int n, List<SimpleClassifier> sc)
Map<Double,AMapping>	getMappings(double min, double max, int n, SimpleClassifier cp) Computes the mappings for a classifier across a dimension of the mesh
ComplexClassifier	getZoomedHillTop(int gridPoints, int iterations, List<SimpleClassifier> sc) Generates the grids iteratively to find the best possible solution
ComplexClassifier	getZoomedHillTop(int gridPoints, long duration, List<SimpleClassifier> sc) Alternative: Generates the grids iteratively to find the best possible solution, within a given maxmimal duration in seconds
void	setStrategy(String strategy)

Methods inherited from class org.aksw.limes.core.ml.algorithm.euclid.BooleanSelfConfigurator

computeNext, getIntersection, getMapping, learnClassifer

Methods inherited from class org.aksw.limes.core.ml.algorithm.euclid.LinearSelfConfigurator

computeMeasure, computeNext, computeQuality, execute, executeClassifier, getAllInitialClassifiers, getBestInitialClassifiers, getBestInitialClassifiers, getBestOneToOneMapping, getInitialOverallClassifiers, getLinkSpecification, getMeasure, getOverallMapping, getPropertyStats, getPropertyType, getResults, getSource, getTarget, getThreshold, minimizeToKnow, normalizeClassifiers, setDefaultMeasures, setMeasure, setMeasure, setPFMType, setSource, setSupervisedBatch, setTarget

Methods inherited from class java.lang.Object

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

Field Detail

MAX_STRATEGY

public static String MAX_STRATEGY

MIN_STRATEFY

public static String MIN_STRATEFY

Constructor Detail

MeshBasedSelfConfigurator

public MeshBasedSelfConfigurator(ACache source,
                                 ACache target)

Basic constructor uses default values for minCoverage and beta.

Parameters:

source -

target -

MeshBasedSelfConfigurator

public MeshBasedSelfConfigurator(ACache source,
                                 ACache target,
                                 double minCoverage)

Constructor

Parameters:

source - Source cache

target - Target cache

minCoverage - Minimal coverage for a property to be considered for linking

Method Detail

copy

public static List<Double> copy(List<Double> l)

generateCoordinates

public static List<List<Double>> generateCoordinates(List<Double> min,
                                                     List<Double> max,
                                                     int n)

Generates a mesh over the space defined by the mins and max set. The size of the mesh is set to ensure that each side of the mesh consists of exactly n points

Parameters:

min - Coordinates on the bottom-left point (in 2D)

max - Coordinates on the top-right point (in 2D)

n - Size of the mesh

Returns:

List of coordinates for the mesh points

getMappings

public Map<Double,AMapping> getMappings(double min,
                                        double max,
                                        int n,
                                        SimpleClassifier cp)

Computes the mappings for a classifier across a dimension of the mesh

Parameters:

min -

max -

n -

cp -

Returns:

getZoomedHillTop

public ComplexClassifier getZoomedHillTop(int gridPoints,
                                          int iterations,
                                          List<SimpleClassifier> sc)

Generates the grids iteratively to find the best possible solution

Parameters:

gridPoints - Number of points used for the grid. 5 leads to a 4x4x..x4 grid

iterations - Number of times that the zooming is applied

sc - Initial classifier, defines the dimensions of the space

Returns:

A complex classifier

getZoomedHillTop

public ComplexClassifier getZoomedHillTop(int gridPoints,
                                          long duration,
                                          List<SimpleClassifier> sc)

Alternative: Generates the grids iteratively to find the best possible solution, within a given maxmimal duration in seconds

Parameters:

gridPoints - Number of points used for the grid. 5 leads to a

duration - maximal duration in seconds

sc - Initial classifier, defines the dimensions of the space

Returns:

A complex classifier

getHillTop

public ComplexClassifier getHillTop(int n,
                                    List<SimpleClassifier> sc)

getHillTop

public ComplexClassifier getHillTop(List<Double> min,
                                    List<Double> max,
                                    int n,
                                    List<SimpleClassifier> sc)

Parameters:

min -

max -

n -

sc -

Returns:

setStrategy

public void setStrategy(String strategy)