Using the JDM API - Java Data Mining: Strategy, Standard, and Practice - page 208

Table 9-9 javax.datamining.MiningFunction enumerations

javax.datamining.data package

Physical data specification related interfaces

PhysicalDataSet A PhysicalDataSet instance describes data for mining. It encapsulates data loca-

tion and access information. It can have a set of physical attributes that describe
the physical nature of the data, such as attribute name, data type, and role.

PhysicalAttribute A PhysicalAttribute instance describes an attribute of a physical data source.

PhysicalDataRecord A PhysicalDataRecord instance represents a set of attribute name-value

pairs and enables singleton record scoring. Users can set specific values for
PhysicalAttribute instances.

Logical data specification related interfaces

LogicalData A LogicalData instance is a set of logical attributes that describes the logical

nature of the data used as input for model building.

LogicalAttribute A LogicalAttribute instance describes the data to be used as input to data

mining operations. As such, a logical attribute references additional metadata
that characterizes the attribute as either categorical or numerical.

CategorySet A CategorySet instance specifies a set of categorical values that constitute a

categorical attribute, cost matrix, or confusion matrix. Each category exists at
an index from 0 to N 1, where N is the cardinality of the set. To have NULL be

a valid value, it must be explicitly assigned or use the default property of
“valid”. To have NULL as a missing value, it must be explicitly identified as
“missing” using the CategoryProperty enumeration.

Data related enumerations

AttributeDataType AttributeDataType enumerates the JDM data types: integer, double, and string.

PhysicalAttributeRole A role defines how an algorithm or mining operation will interpret the values of

an attribute. PhysicalAttributeRole enumerations include case id, transactional
data attribute roles, and taxonomy-related attribute roles.

AttributeType AttributeType enumerates the options for indicating how an attribute is to be

interpreted by mining algorithms. For example, an attribute whose values are
integers may be treated as categorical instead of the default numerical.

DataPreparationStatus DataPreparationStatus enumerates the data preprocessing options for an

attribute: either prepared or unprepared. For example, to indicate that the
attribute needs no further preprocessing steps, the user can specify the attribute
as prepared.

CategoryProperty CategoryProperty enumerates the types of categorical attribute values, i.e.,

categories. For example, it can be used to specify whether the attribute NULL
values are treated as valid, invalid, missing, or error.