|
![]() |
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectde.lmu.ifi.dbs.elki.algorithm.AbstractAlgorithm<R>
de.lmu.ifi.dbs.elki.algorithm.clustering.AbstractProjectedClustering<Clustering<Model>,V>
de.lmu.ifi.dbs.elki.algorithm.clustering.correlation.ORCLUS<V>
V
- the type of NumberVector handled by this Algorithm@Title(value="ORCLUS: Arbitrarily ORiented projected CLUSter generation") @Description(value="Algorithm to find correlation clusters in high dimensional spaces.") @Reference(authors="C. C. Aggarwal, P. S. Yu", title="Finding Generalized Projected Clusters in High Dimensional Spaces", booktitle="Proc. ACM SIGMOD Int. Conf. on Management of Data (SIGMOD \'00)", url="http://dx.doi.org/10.1145/342009.335383") public class ORCLUS<V extends NumberVector<V,?>>
ORCLUS provides the ORCLUS algorithm, an algorithm to find clusters in high dimensional spaces.
Reference: C. C. Aggarwal, P. S. Yu: Finding Generalized Projected Clusters
in High Dimensional Spaces.
In: Proc. ACM SIGMOD Int. Conf. on Management of Data (SIGMOD '00).
Nested Class Summary | |
---|---|
private class |
ORCLUS.ORCLUSCluster
Encapsulates the attributes of a cluster. |
static class |
ORCLUS.Parameterizer<V extends NumberVector<V,?>>
Parameterization class. |
private class |
ORCLUS.ProjectedEnergy
Encapsulates the projected energy for a cluster. |
Field Summary | |
---|---|
private double |
alpha
Holds the value of ALPHA_ID . |
static OptionID |
ALPHA_ID
Parameter to specify the factor for reducing the number of current clusters in each iteration, must be an integer greater than 0 and less than 1. |
private static Logging |
logger
The logger for this class. |
private PCARunner<V> |
pca
The PCA utility object. |
private Long |
seed
Holds the value of SEED_ID . |
static OptionID |
SEED_ID
Parameter to specify the random generator seed. |
Fields inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.AbstractProjectedClustering |
---|
k, k_i, K_I_ID, K_ID, l, L_ID |
Constructor Summary | |
---|---|
ORCLUS(int k,
int k_i,
int l,
double alpha,
long seed,
PCARunner<V> pca)
Java constructor. |
Method Summary | |
---|---|
private void |
assign(Relation<V> database,
DistanceQuery<V,DoubleDistance> distFunc,
List<ORCLUS.ORCLUSCluster> clusters)
Creates a partitioning of the database by assigning each object to its closest seed. |
private Matrix |
findBasis(Relation<V> database,
DistanceQuery<V,DoubleDistance> distFunc,
ORCLUS.ORCLUSCluster cluster,
int dim)
Finds the basis of the subspace of dimensionality dim for the
specified cluster. |
TypeInformation[] |
getInputTypeRestriction()
Get the input type restriction used for negotiating the data query. |
protected Logging |
getLogger()
Get the (STATIC) logger for this class. |
private List<ORCLUS.ORCLUSCluster> |
initialSeeds(Relation<V> database,
int k)
Initializes the list of seeds wit a random sample of size k. |
private void |
merge(Relation<V> database,
DistanceQuery<V,DoubleDistance> distFunc,
List<ORCLUS.ORCLUSCluster> clusters,
int k_new,
int d_new,
IndefiniteProgress cprogress)
Reduces the number of seeds to k_new |
private ORCLUS.ProjectedEnergy |
projectedEnergy(Relation<V> database,
DistanceQuery<V,DoubleDistance> distFunc,
ORCLUS.ORCLUSCluster c_i,
ORCLUS.ORCLUSCluster c_j,
int i,
int j,
int dim)
Computes the projected energy of the specified clusters. |
private V |
projection(ORCLUS.ORCLUSCluster c,
V o,
V factory)
Returns the projection of real vector o in the subspace of cluster c. |
Clustering<Model> |
run(Database database,
Relation<V> relation)
Performs the ORCLUS algorithm on the given database. |
private ORCLUS.ORCLUSCluster |
union(Relation<V> database,
DistanceQuery<V,DoubleDistance> distFunc,
ORCLUS.ORCLUSCluster c1,
ORCLUS.ORCLUSCluster c2,
int dim)
Returns the union of the two specified clusters. |
Methods inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.AbstractProjectedClustering |
---|
getDistanceFunction, getDistanceQuery |
Methods inherited from class de.lmu.ifi.dbs.elki.algorithm.AbstractAlgorithm |
---|
makeParameterDistanceFunction, run |
Methods inherited from class java.lang.Object |
---|
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Methods inherited from interface de.lmu.ifi.dbs.elki.algorithm.clustering.ClusteringAlgorithm |
---|
run |
Field Detail |
---|
private static final Logging logger
public static final OptionID ALPHA_ID
Default value: 0.5
Key: -orclus.alpha
public static final OptionID SEED_ID
private double alpha
ALPHA_ID
.
private Long seed
SEED_ID
.
private PCARunner<V extends NumberVector<V,?>> pca
Constructor Detail |
---|
public ORCLUS(int k, int k_i, int l, double alpha, long seed, PCARunner<V> pca)
k
- k Parameterk_i
- k_i Parameterl
- l Parameteralpha
- Alpha Parameterseed
- Seed parameterpca
- PCA runnerMethod Detail |
---|
public Clustering<Model> run(Database database, Relation<V> relation) throws IllegalStateException
IllegalStateException
private List<ORCLUS.ORCLUSCluster> initialSeeds(Relation<V> database, int k)
database
- the database holding the objectsk
- the size of the random sample
private void assign(Relation<V> database, DistanceQuery<V,DoubleDistance> distFunc, List<ORCLUS.ORCLUSCluster> clusters)
database
- the database holding the objectsdistFunc
- distance functionclusters
- the array of clusters to which the objects should be
assigned toprivate Matrix findBasis(Relation<V> database, DistanceQuery<V,DoubleDistance> distFunc, ORCLUS.ORCLUSCluster cluster, int dim)
dim
for the
specified cluster.
database
- the database to run the algorithm ondistFunc
- the distance functioncluster
- the clusterdim
- the dimensionality of the subspace
private void merge(Relation<V> database, DistanceQuery<V,DoubleDistance> distFunc, List<ORCLUS.ORCLUSCluster> clusters, int k_new, int d_new, IndefiniteProgress cprogress)
database
- the database holding the objectsdistFunc
- the distance functionclusters
- the set of current seedsk_new
- the new number of seedsd_new
- the new dimensionality of the subspaces for each seedprivate ORCLUS.ProjectedEnergy projectedEnergy(Relation<V> database, DistanceQuery<V,DoubleDistance> distFunc, ORCLUS.ORCLUSCluster c_i, ORCLUS.ORCLUSCluster c_j, int i, int j, int dim)
database
- the database holding the objectsdistFunc
- the distance functionc_i
- the first clusterc_j
- the second clusteri
- the index of cluster c_i in the cluster listj
- the index of cluster c_j in the cluster listdim
- the dimensionality of the clusters
private ORCLUS.ORCLUSCluster union(Relation<V> database, DistanceQuery<V,DoubleDistance> distFunc, ORCLUS.ORCLUSCluster c1, ORCLUS.ORCLUSCluster c2, int dim)
database
- the database holding the objectsdistFunc
- the distance functionc1
- the first clusterc2
- the second clusterdim
- the dimensionality of the union cluster
private V projection(ORCLUS.ORCLUSCluster c, V o, V factory)
c
- the clustero
- the double vectorfactory
- Factory object / prototype
public TypeInformation[] getInputTypeRestriction()
AbstractAlgorithm
getInputTypeRestriction
in interface Algorithm
getInputTypeRestriction
in class AbstractAlgorithm<Clustering<Model>>
protected Logging getLogger()
AbstractAlgorithm
getLogger
in class AbstractAlgorithm<Clustering<Model>>
|
|
|||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |