de.lmu.ifi.dbs.elki.evaluation.roc

Class ROC

java.lang.Object

de.lmu.ifi.dbs.elki.evaluation.roc.ROC

public class ROC
extends Object

Compute ROC (Receiver Operating Characteristics) curves. A ROC curve compares the true positive rate (y-axis) and false positive rate (x-axis). It was first used in radio signal detection, but has since found widespread use in information retrieval, in particular for evaluating binary classification problems. ROC curves are particularly useful to evaluate a ranking of objects with respect to a binary classification problem: a random sampling will approximately achieve a ROC value of 0.5, while a perfect separation will achieve 1.0 (all positives first) or 0.0 (all negatives first). In most use cases, a score significantly below 0.5 indicates that the algorithm result has been used the wrong way, and should be used backwards.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	ROC.DistanceResultAdapter<D extends Distance<D>> This adapter can be used for an arbitrary collection of Integers, and uses that id1.compareTo(id2) !
static class	ROC.OutlierScoreAdapter This adapter can be used for an arbitrary collection of Integers, and uses that id1.compareTo(id2) !
static class	ROC.SimpleAdapter This adapter can be used for an arbitrary collection of Integers, and uses that id1.compareTo(id2) !

Constructor Summary

Constructors

Constructor and Description
ROC()

Method Summary

Methods

Modifier and Type	Method and Description
static <D extends Distance<D>> double	computeROCAUCDistanceResult(int size, Cluster<?> clus, DistanceDBIDResult<D> nei) Compute a ROC curves Area-under-curve for a QueryResult and a Cluster.
static <D extends Distance<D>> double	computeROCAUCDistanceResult(int size, DBIDs ids, DistanceDBIDResult<D> nei) Compute a ROC curves Area-under-curve for a QueryResult and a Cluster.
static double	computeROCAUCSimple(int size, DBIDs ids, DBIDs nei) Compute a ROC curves Area-under-curve for a QueryResult and a Cluster.
static <C extends Comparable<? super C>,T> XYCurve	materializeROC(int size, Set<? super T> ids, Iterator<? extends PairInterface<C,T>> nei) Compute a ROC curve given a set of positive IDs and a sorted list of (comparable, ID)s, where the comparable object is used to decided when two objects are interchangeable.
static <C extends Comparable<? super C>> XYCurve	materializeROC(int size, SetDBIDs ids, Iterator<? extends PairInterface<C,? extends DBIDRef>> nei) Compute a ROC curve given a set of positive IDs and a sorted list of (comparable, ID)s, where the comparable object is used to decided when two objects are interchangeable.

Methods inherited from class java.lang.Object

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

Constructor Detail

ROC

public ROC()

Method Detail

materializeROC

public static <C extends Comparable<? super C>,T> XYCurve materializeROC(int size,
                                                         Set<? super T> ids,
                                                         Iterator<? extends PairInterface<C,T>> nei)

Compute a ROC curve given a set of positive IDs and a sorted list of (comparable, ID)s, where the comparable object is used to decided when two objects are interchangeable.

Type Parameters:

C - Reference type

Parameters:

size - Database size

ids - Collection of positive IDs, should support efficient contains()

nei - List of neighbors along with some comparable object to detect 'same positions'.

Returns:

area under curve

materializeROC

public static <C extends Comparable<? super C>> XYCurve materializeROC(int size,
                                                       SetDBIDs ids,
                                                       Iterator<? extends PairInterface<C,? extends DBIDRef>> nei)

Compute a ROC curve given a set of positive IDs and a sorted list of (comparable, ID)s, where the comparable object is used to decided when two objects are interchangeable.

Type Parameters:

C - Reference type

Parameters:

size - Database size

ids - Collection of positive IDs, should support efficient contains()

nei - List of neighbors along with some comparable object to detect 'same positions'.

Returns:

area under curve

computeROCAUCDistanceResult

public static <D extends Distance<D>> double computeROCAUCDistanceResult(int size,
                                                         Cluster<?> clus,
                                                         DistanceDBIDResult<D> nei)

Compute a ROC curves Area-under-curve for a QueryResult and a Cluster.

Type Parameters:

D - Distance type

Parameters:

size - Database size

clus - Cluster object

nei - Query result

Returns:

area under curve

computeROCAUCDistanceResult

public static <D extends Distance<D>> double computeROCAUCDistanceResult(int size,
                                                         DBIDs ids,
                                                         DistanceDBIDResult<D> nei)

Compute a ROC curves Area-under-curve for a QueryResult and a Cluster.

Type Parameters:

D - Distance type

Parameters:

size - Database size

ids - Collection of positive IDs, should support efficient contains()

nei - Query Result

Returns:

area under curve

computeROCAUCSimple

public static double computeROCAUCSimple(int size,
                         DBIDs ids,
                         DBIDs nei)

Compute a ROC curves Area-under-curve for a QueryResult and a Cluster.

Parameters:

size - Database size

ids - Collection of positive IDs, should support efficient contains()

nei - Query Result

Returns:

area under curve