Main Page | Class Hierarchy | Class List | Class Members

kddml.Operators.Postprocessing.CLUSTER_NUMBER_RESOLVER Class Reference

Inheritance diagram for kddml.Operators.Postprocessing.CLUSTER_NUMBER_RESOLVER:

kddml.Operators.Postprocessing.PostProcessingResolver kddml.Operators.OperatorResolver kddml.Operators.HWResourcesDescription List of all members.

Public Member Functions

KDDMLObject execute (Hashtable< String, KDDMLScalarManager > settings, Vector< KDDMLObject > arguments) throws ResolverException, KDDMLCoreException
KDDMLObjectType getArgumentType (int i)
boolean highDiskAccessIsNeeded ()

Detailed Description

Given a cluster model and a dataset, this operator returns only the tuples of the dataset belonging to a specified cluster. The mapping between attributes used in the clustering and attributes in the input table is by name. Therefore, the clustering model and the data source must be compatible. In particular, a table is compatible with a model if for each active mining field belonging to the model, there is an attribute in the table with the same name and type. The prediction of which cluster contains the instance is performed by comparing the instance and the cluster centroid. When two records are compared then either the distance or the similarity is of interest. In both cases the measures can be computed by a combination of an inner function and an outer function. The inner function compares two single fields (e.g. absolute difference between continuous attributes) values and the outer function (e.g. euclidean distance) computes an aggregation over all fields. Inner function and outer function are stored in the clustering model when the clusters are built. The operator can be also used to predict the tuples belonging to the cluster with maximal cardinality.

Title: KDDML

Description: Knowledge Discovery in Database Environment

Copyright: Copyright (c) 2003 - 2005

Company: Universita' di Pisa - Dipartimento di Informatica

Author:
Andrea Romei (romei@di.unipi.it)
Version:
2.0.16


Member Function Documentation

KDDMLObject kddml.Operators.Postprocessing.CLUSTER_NUMBER_RESOLVER.execute Hashtable< String, KDDMLScalarManager settings,
Vector< KDDMLObject arguments
throws ResolverException, KDDMLCoreException [virtual]
 

Returns the result of evaluating the operator over the passed arguments and attributes.
Attributes are given as hashtable, where the key is the name of the XML attribute related to the operator and the value is a KDDMLScalar object containing the attribute value. Attribute value is checked by the interpreter layer and it is correct.
Arguments are given as vector of KDDMLOjbect. Types of arguments are correct, as required by the getArgumentType(int i) method.
Implementations of the execute method can be distinguished according to the type of operator.

Parameters:
settings Hashtable the attributes related to the operator. The key of the hashtable is the name of the attribute. The value of the hashtable is a KDDMLScalar representing the value of the attribute.
arguments Vector the list of arguments as KDDMLObject.
Returns:
KDDMLObject the result of evaluating the operator.
Exceptions:
ResolverException if a resolving error occurs.
KDDMLCoreException if a level core error occurs.

Implements kddml.Operators.OperatorResolver.

KDDMLObjectType kddml.Operators.Postprocessing.CLUSTER_NUMBER_RESOLVER.getArgumentType int  i  )  [virtual]
 

Returns the expected type of the i-th argument of the operator. Returns null if the operator do not admit argument as i-th child.

Parameters:
i int
Returns:
KDDMLObjectType

Implements kddml.Operators.OperatorResolver.

boolean kddml.Operators.Postprocessing.CLUSTER_NUMBER_RESOLVER.highDiskAccessIsNeeded  ) 
 

Returns true if the operator performe a high number of I/O operation to the hard disk.

Returns:
boolean

Reimplemented from kddml.Operators.OperatorResolver.


Generated on Thu Feb 23 13:04:50 2006 for kddml by  doxygen 1.4.3