Package oracle.dmt.odm.data

This package contains classes that are used to describe the input to data mining operations.

See:
Description

Class Summary

Attribute	An instance of Attribute maps to a column with a name and datatype.
AttributeInstance	An instance of AttributeInstance supports providing named and typed values as input for various functions.
LogicalDataSpecification	An instance of LogicalDataSpecification (LDS) is used to describe the logical characteristics of the data used in model building, and is composed of a set of mining attributes.
MiningAttribute	An instance of MiningAttribute is a logical concept that describes a domain of data to be used as input to data mining operations.
ModelSignature	An instance of ModelSignature specifies the input attributes required to apply data using a specific model.
NonTransactionalDataSpecification	A NonTransactionalDataSpecification instance instructs the DMS to treat associated data as "non-transactional", that is, the data consists of one record (row) per case.
PhysicalDataSpecification	The abstract class PhysicalDataSpecification is the common superclass for all data format classes.
SignatureAttribute	An instance of SignatureAttribute is used to represent a logical attribute that was used for model building and is required for scoring data using a specific model.
TransactionalDataSpecification	An instance of TransactionalDataSpecification instructs the DMS to treat associated data as transactional, i.e., a given case is stored in multiple records in a table with column roles: sequenceID, attributeName, and value.

Package oracle.dmt.odm.data Description

This package contains classes that are used to describe the input to data mining operations. Physical data is mapped to Logical data used for mining operations. Oracle Data Mining supports data in two formats:

• Non-transactional data: This corresponds to traditional two-dimensional tables. Typically, each column is mapped to a MiningAttribute. The number of columns cannot exceed 1000. Non-transactional data is also referred to as "single record case" format.
• Transactional data: Data in this format is represented in three columns: sequence ID, attribute name, and value. This is a pivoted representation of non-transactional data. The sequence ID associates the cells of a row in the non-transactional data together. Transactional data is also referred to as "multi-record case" format.