Search in sources :

Example 96 with OutlierResult

use of de.lmu.ifi.dbs.elki.result.outlier.OutlierResult in project elki by elki-project.

the class TrivialNoOutlier method run.

/**
 * Run the actual algorithm.
 *
 * @param relation Relation
 * @return Result
 */
public OutlierResult run(Relation<?> relation) {
    WritableDoubleDataStore scores = DataStoreUtil.makeDoubleStorage(relation.getDBIDs(), DataStoreFactory.HINT_HOT);
    for (DBIDIter iditer = relation.iterDBIDs(); iditer.valid(); iditer.advance()) {
        scores.putDouble(iditer, 0.0);
    }
    DoubleRelation scoreres = new MaterializedDoubleRelation("Trivial no-outlier score", "no-outlier", scores, relation.getDBIDs());
    OutlierScoreMeta meta = new ProbabilisticOutlierScore();
    return new OutlierResult(meta, scoreres);
}
Also used : WritableDoubleDataStore(de.lmu.ifi.dbs.elki.database.datastore.WritableDoubleDataStore) OutlierResult(de.lmu.ifi.dbs.elki.result.outlier.OutlierResult) ProbabilisticOutlierScore(de.lmu.ifi.dbs.elki.result.outlier.ProbabilisticOutlierScore) DoubleRelation(de.lmu.ifi.dbs.elki.database.relation.DoubleRelation) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation) OutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.OutlierScoreMeta) DBIDIter(de.lmu.ifi.dbs.elki.database.ids.DBIDIter)

Example 97 with OutlierResult

use of de.lmu.ifi.dbs.elki.result.outlier.OutlierResult in project elki by elki-project.

the class ComputeOutlierHistogram method processNewResult.

@Override
public void processNewResult(ResultHierarchy hier, Result newResult) {
    final Database db = ResultUtil.findDatabase(hier);
    List<OutlierResult> ors = ResultUtil.filterResults(hier, newResult, OutlierResult.class);
    if (ors == null || ors.isEmpty()) {
        // "+ComputeOutlierHistogram.class.getSimpleName());
        return;
    }
    for (OutlierResult or : ors) {
        db.getHierarchy().add(or, evaluateOutlierResult(db, or));
    }
}
Also used : Database(de.lmu.ifi.dbs.elki.database.Database) OutlierResult(de.lmu.ifi.dbs.elki.result.outlier.OutlierResult)

Example 98 with OutlierResult

use of de.lmu.ifi.dbs.elki.result.outlier.OutlierResult in project elki by elki-project.

the class SLOM method run.

/**
 * @param database Database to process
 * @param spatial Spatial Relation to use.
 * @param relation Relation to use.
 * @return Outlier detection result
 */
public OutlierResult run(Database database, Relation<N> spatial, Relation<O> relation) {
    final NeighborSetPredicate npred = getNeighborSetPredicateFactory().instantiate(database, spatial);
    DistanceQuery<O> distFunc = getNonSpatialDistanceFunction().instantiate(relation);
    WritableDoubleDataStore modifiedDistance = DataStoreUtil.makeDoubleStorage(relation.getDBIDs(), DataStoreFactory.HINT_HOT | DataStoreFactory.HINT_TEMP);
    // calculate D-Tilde
    for (DBIDIter iditer = relation.iterDBIDs(); iditer.valid(); iditer.advance()) {
        double sum = 0;
        double maxDist = 0;
        int cnt = 0;
        final DBIDs neighbors = npred.getNeighborDBIDs(iditer);
        for (DBIDIter iter = neighbors.iter(); iter.valid(); iter.advance()) {
            if (DBIDUtil.equal(iditer, iter)) {
                continue;
            }
            double dist = distFunc.distance(iditer, iter);
            sum += dist;
            cnt++;
            maxDist = Math.max(maxDist, dist);
        }
        if (cnt > 1) {
            modifiedDistance.putDouble(iditer, ((sum - maxDist) / (cnt - 1)));
        } else {
            // Use regular distance when the d-tilde trick is undefined.
            // Note: this can be 0 when there were no neighbors.
            modifiedDistance.putDouble(iditer, maxDist);
        }
    }
    // Second step - compute actual SLOM values
    DoubleMinMax slomminmax = new DoubleMinMax();
    WritableDoubleDataStore sloms = DataStoreUtil.makeDoubleStorage(relation.getDBIDs(), DataStoreFactory.HINT_STATIC);
    for (DBIDIter iditer = relation.iterDBIDs(); iditer.valid(); iditer.advance()) {
        double sum = 0;
        int cnt = 0;
        final DBIDs neighbors = npred.getNeighborDBIDs(iditer);
        for (DBIDIter iter = neighbors.iter(); iter.valid(); iter.advance()) {
            if (DBIDUtil.equal(iditer, iter)) {
                continue;
            }
            sum += modifiedDistance.doubleValue(iter);
            cnt++;
        }
        double slom;
        if (cnt > 0) {
            // With and without the object itself:
            double avgPlus = (sum + modifiedDistance.doubleValue(iditer)) / (cnt + 1);
            double avg = sum / cnt;
            double beta = 0;
            for (DBIDIter iter = neighbors.iter(); iter.valid(); iter.advance()) {
                final double dist = modifiedDistance.doubleValue(iter);
                if (dist > avgPlus) {
                    beta += 1;
                } else if (dist < avgPlus) {
                    beta -= 1;
                }
            }
            // Include object itself
            if (!neighbors.contains(iditer)) {
                final double dist = modifiedDistance.doubleValue(iditer);
                if (dist > avgPlus) {
                    beta += 1;
                } else if (dist < avgPlus) {
                    beta -= 1;
                }
            }
            beta = Math.abs(beta);
            // note: cnt == size of N(x), not N+(x)
            if (cnt > 1) {
                beta = Math.max(beta, 1.0) / (cnt - 1);
            } else {
                // Workaround insufficiency in SLOM paper - div by zero
                beta = 1.0;
            }
            beta = beta / (1 + avg);
            slom = beta * modifiedDistance.doubleValue(iditer);
        } else {
            // No neighbors to compare to - no score.
            slom = 0.0;
        }
        sloms.putDouble(iditer, slom);
        slomminmax.put(slom);
    }
    DoubleRelation scoreResult = new MaterializedDoubleRelation("SLOM", "slom-outlier", sloms, relation.getDBIDs());
    OutlierScoreMeta scoreMeta = new BasicOutlierScoreMeta(slomminmax.getMin(), slomminmax.getMax(), 0.0, Double.POSITIVE_INFINITY);
    OutlierResult or = new OutlierResult(scoreMeta, scoreResult);
    or.addChildResult(npred);
    return or;
}
Also used : WritableDoubleDataStore(de.lmu.ifi.dbs.elki.database.datastore.WritableDoubleDataStore) DBIDs(de.lmu.ifi.dbs.elki.database.ids.DBIDs) OutlierResult(de.lmu.ifi.dbs.elki.result.outlier.OutlierResult) NeighborSetPredicate(de.lmu.ifi.dbs.elki.algorithm.outlier.spatial.neighborhood.NeighborSetPredicate) DoubleRelation(de.lmu.ifi.dbs.elki.database.relation.DoubleRelation) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation) BasicOutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.BasicOutlierScoreMeta) OutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.OutlierScoreMeta) BasicOutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.BasicOutlierScoreMeta) DBIDIter(de.lmu.ifi.dbs.elki.database.ids.DBIDIter) DoubleMinMax(de.lmu.ifi.dbs.elki.math.DoubleMinMax) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation)

Example 99 with OutlierResult

use of de.lmu.ifi.dbs.elki.result.outlier.OutlierResult in project elki by elki-project.

the class FastABOD method run.

/**
 * Run Fast-ABOD on the data set.
 *
 * @param relation Relation to process
 * @return Outlier detection result
 */
@Override
public OutlierResult run(Database db, Relation<V> relation) {
    DBIDs ids = relation.getDBIDs();
    // Build a kernel matrix, to make O(n^3) slightly less bad.
    SimilarityQuery<V> sq = db.getSimilarityQuery(relation, kernelFunction);
    KernelMatrix kernelMatrix = new KernelMatrix(sq, relation, ids);
    WritableDoubleDataStore abodvalues = DataStoreUtil.makeDoubleStorage(ids, DataStoreFactory.HINT_STATIC);
    DoubleMinMax minmaxabod = new DoubleMinMax();
    MeanVariance s = new MeanVariance();
    KNNHeap nn = DBIDUtil.newHeap(k);
    for (DBIDIter pA = ids.iter(); pA.valid(); pA.advance()) {
        final double simAA = kernelMatrix.getSimilarity(pA, pA);
        // Choose the k-min nearest
        nn.clear();
        for (DBIDIter nB = relation.iterDBIDs(); nB.valid(); nB.advance()) {
            if (DBIDUtil.equal(nB, pA)) {
                continue;
            }
            double simBB = kernelMatrix.getSimilarity(nB, nB);
            double simAB = kernelMatrix.getSimilarity(pA, nB);
            double sqdAB = simAA + simBB - simAB - simAB;
            if (!(sqdAB > 0.)) {
                continue;
            }
            nn.insert(sqdAB, nB);
        }
        KNNList nl = nn.toKNNList();
        s.reset();
        DoubleDBIDListIter iB = nl.iter(), iC = nl.iter();
        for (; iB.valid(); iB.advance()) {
            double sqdAB = iB.doubleValue();
            double simAB = kernelMatrix.getSimilarity(pA, iB);
            if (!(sqdAB > 0.)) {
                continue;
            }
            for (iC.seek(iB.getOffset() + 1); iC.valid(); iC.advance()) {
                double sqdAC = iC.doubleValue();
                double simAC = kernelMatrix.getSimilarity(pA, iC);
                if (!(sqdAC > 0.)) {
                    continue;
                }
                // Exploit bilinearity of scalar product:
                // <B-A, C-A> = <B, C-A> - <A,C-A>
                // = <B,C> - <B,A> - <A,C> + <A,A>
                double simBC = kernelMatrix.getSimilarity(iB, iC);
                double numerator = simBC - simAB - simAC + simAA;
                double div = 1. / (sqdAB * sqdAC);
                s.put(numerator * div, FastMath.sqrt(div));
            }
        }
        // Sample variance probably would probably be better, but the ABOD
        // publication uses the naive variance.
        final double abof = s.getNaiveVariance();
        minmaxabod.put(abof);
        abodvalues.putDouble(pA, abof);
    }
    // Build result representation.
    DoubleRelation scoreResult = new MaterializedDoubleRelation("Angle-Based Outlier Degree", "abod-outlier", abodvalues, relation.getDBIDs());
    OutlierScoreMeta scoreMeta = new InvertedOutlierScoreMeta(minmaxabod.getMin(), minmaxabod.getMax(), 0.0, Double.POSITIVE_INFINITY);
    return new OutlierResult(scoreMeta, scoreResult);
}
Also used : DoubleDBIDListIter(de.lmu.ifi.dbs.elki.database.ids.DoubleDBIDListIter) WritableDoubleDataStore(de.lmu.ifi.dbs.elki.database.datastore.WritableDoubleDataStore) DBIDs(de.lmu.ifi.dbs.elki.database.ids.DBIDs) OutlierResult(de.lmu.ifi.dbs.elki.result.outlier.OutlierResult) InvertedOutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.InvertedOutlierScoreMeta) KNNHeap(de.lmu.ifi.dbs.elki.database.ids.KNNHeap) DoubleRelation(de.lmu.ifi.dbs.elki.database.relation.DoubleRelation) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation) OutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.OutlierScoreMeta) InvertedOutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.InvertedOutlierScoreMeta) DBIDIter(de.lmu.ifi.dbs.elki.database.ids.DBIDIter) KernelMatrix(de.lmu.ifi.dbs.elki.distance.similarityfunction.kernel.KernelMatrix) MeanVariance(de.lmu.ifi.dbs.elki.math.MeanVariance) DoubleMinMax(de.lmu.ifi.dbs.elki.math.DoubleMinMax) KNNList(de.lmu.ifi.dbs.elki.database.ids.KNNList) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation)

Example 100 with OutlierResult

use of de.lmu.ifi.dbs.elki.result.outlier.OutlierResult in project elki by elki-project.

the class ISOS method run.

/**
 * Run the algorithm.
 *
 * @param relation data relation.
 * @return outlier detection result
 */
public OutlierResult run(Relation<O> relation) {
    // Query size
    final int k1 = k + 1;
    final double perplexity = k / 3.;
    KNNQuery<O> knnq = relation.getKNNQuery(getDistanceFunction(), k1);
    final double logPerp = FastMath.log(perplexity);
    double[] p = new double[k + 10];
    ModifiableDoubleDBIDList dists = DBIDUtil.newDistanceDBIDList(k + 10);
    DoubleDBIDListIter di = dists.iter();
    FiniteProgress prog = LOG.isVerbose() ? new FiniteProgress("ISOS scores", relation.size(), LOG) : null;
    WritableDoubleDataStore scores = DataStoreUtil.makeDoubleStorage(relation.getDBIDs(), DataStoreFactory.HINT_HOT | DataStoreFactory.HINT_DB, 1.);
    for (DBIDIter it = relation.iterDBIDs(); it.valid(); it.advance()) {
        KNNList knns = knnq.getKNNForDBID(it, k1);
        if (p.length < knns.size() + 1) {
            p = new double[knns.size() + 10];
        }
        final DoubleDBIDListIter ki = knns.iter();
        try {
            double id = estimateID(it, ki, p);
            adjustDistances(it, ki, knns.getKNNDistance(), id, dists);
            // We now continue with the modified distances:
            // Compute affinities
            SOS.computePi(it, di, p, perplexity, logPerp);
            // Normalization factor:
            double s = SOS.sumOfProbabilities(it, di, p);
            if (s > 0) {
                nominateNeighbors(it, di, p, 1. / s, scores);
            }
        } catch (ArithmeticException e) {
            // ID estimation failed, supposedly constant values.
            if (knns.size() > 1) {
                Arrays.fill(p, 1. / (knns.size() - 1));
                nominateNeighbors(it, di, p, 1., scores);
            }
        }
        LOG.incrementProcessed(prog);
    }
    LOG.ensureCompleted(prog);
    DoubleMinMax minmax = transformScores(scores, relation.getDBIDs(), logPerp, phi);
    DoubleRelation scoreres = new MaterializedDoubleRelation("Intrinsic Stoachastic Outlier Selection", "isos-outlier", scores, relation.getDBIDs());
    OutlierScoreMeta meta = new ProbabilisticOutlierScore(minmax.getMin(), minmax.getMax(), 0.);
    return new OutlierResult(meta, scoreres);
}
Also used : WritableDoubleDataStore(de.lmu.ifi.dbs.elki.database.datastore.WritableDoubleDataStore) FiniteProgress(de.lmu.ifi.dbs.elki.logging.progress.FiniteProgress) OutlierResult(de.lmu.ifi.dbs.elki.result.outlier.OutlierResult) ProbabilisticOutlierScore(de.lmu.ifi.dbs.elki.result.outlier.ProbabilisticOutlierScore) DoubleRelation(de.lmu.ifi.dbs.elki.database.relation.DoubleRelation) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation) GreaterEqualConstraint(de.lmu.ifi.dbs.elki.utilities.optionhandling.constraints.GreaterEqualConstraint) OutlierScoreMeta(de.lmu.ifi.dbs.elki.result.outlier.OutlierScoreMeta) DoubleMinMax(de.lmu.ifi.dbs.elki.math.DoubleMinMax) MaterializedDoubleRelation(de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation)

Aggregations

OutlierResult (de.lmu.ifi.dbs.elki.result.outlier.OutlierResult)144 MaterializedDoubleRelation (de.lmu.ifi.dbs.elki.database.relation.MaterializedDoubleRelation)72 OutlierScoreMeta (de.lmu.ifi.dbs.elki.result.outlier.OutlierScoreMeta)72 WritableDoubleDataStore (de.lmu.ifi.dbs.elki.database.datastore.WritableDoubleDataStore)71 DoubleRelation (de.lmu.ifi.dbs.elki.database.relation.DoubleRelation)71 Database (de.lmu.ifi.dbs.elki.database.Database)69 DoubleMinMax (de.lmu.ifi.dbs.elki.math.DoubleMinMax)62 Test (org.junit.Test)58 DBIDIter (de.lmu.ifi.dbs.elki.database.ids.DBIDIter)57 AbstractOutlierAlgorithmTest (de.lmu.ifi.dbs.elki.algorithm.outlier.AbstractOutlierAlgorithmTest)50 DBIDs (de.lmu.ifi.dbs.elki.database.ids.DBIDs)45 BasicOutlierScoreMeta (de.lmu.ifi.dbs.elki.result.outlier.BasicOutlierScoreMeta)35 ELKIBuilder (de.lmu.ifi.dbs.elki.utilities.ELKIBuilder)26 FiniteProgress (de.lmu.ifi.dbs.elki.logging.progress.FiniteProgress)23 DoubleVector (de.lmu.ifi.dbs.elki.data.DoubleVector)22 KNNList (de.lmu.ifi.dbs.elki.database.ids.KNNList)18 InvertedOutlierScoreMeta (de.lmu.ifi.dbs.elki.result.outlier.InvertedOutlierScoreMeta)13 ProbabilisticOutlierScore (de.lmu.ifi.dbs.elki.result.outlier.ProbabilisticOutlierScore)13 QuotientOutlierScoreMeta (de.lmu.ifi.dbs.elki.result.outlier.QuotientOutlierScoreMeta)13 DoubleDBIDListIter (de.lmu.ifi.dbs.elki.database.ids.DoubleDBIDListIter)11