/*
 * Copyright (C) 2018 Edward Raff
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see .
 */
package jsat.outlier;

import java.util.Arrays;
import jsat.DataSet;
import jsat.classifiers.DataPoint;
import jsat.distributions.multivariate.MultivariateDistribution;
import jsat.distributions.multivariate.NormalMR;
import jsat.utils.concurrent.ParallelUtils;

/**
 * This class provides an outlier detector based upon density estimation.
 *
 * @author Edward Raff
 */
public class DensityOutlier implements Outlier
{
    private double outlierFraction;
    private MultivariateDistribution density;
    /**
     * Threshold for what counts as an outlier. Any value less than or equal to
     * this threshold will be considered an outlier.
     */
    private double threshold;
    
    public DensityOutlier()
    {
        this(0.05);
    }
    
    public DensityOutlier(double outlierFraction)
    {
        this(outlierFraction, new NormalMR());
    }

    public DensityOutlier(double outlierFraction, MultivariateDistribution density)
    {
        this.outlierFraction = outlierFraction;
        this.density = density;
    }
    
    public DensityOutlier(DensityOutlier toCopy)
    {
        this(toCopy.outlierFraction, toCopy.density.clone());
        this.threshold = toCopy.threshold;
    }

    public void setOutlierFraction(double outlierFraction)
    {
        this.outlierFraction = outlierFraction;
    }

    public double getOutlierFraction()
    {
        return outlierFraction;
    }

    public void setDensityDistribution(MultivariateDistribution density)
    {
        this.density = density;
    }

    public MultivariateDistribution getDensityDistribution()
    {
        return density;
    }
    
    
    
    @Override
    public void fit(DataSet d, boolean parallel)
    {
        density.setUsingData(d, parallel);
        double[] scores = new double[d.getSampleSize()];
        ParallelUtils.run(parallel, scores.length, (start, end)->
        {
            for(int i = start; i < end; i++)
                scores[i] = density.logPdf(d.getDataPoint(i).getNumericalValues());
        });
        Arrays.sort(scores);
        threshold = scores[(int)(scores.length*outlierFraction)];
    }

    @Override
    public double score(DataPoint x)
    {
        double logPDF = density.logPdf(x.getNumericalValues());
        return logPDF - threshold;
    }

    @Override
    protected DensityOutlier clone() throws CloneNotSupportedException
    {
        return new DensityOutlier(this);
    }
    
}

 

Ads help maintain this website.