The database objects that are provided by Intelligent Miner® reflect
the current standard for data mining in the context of SQL.
For a code sample that shows how you can build a data mining model,
see SQL code examples for data mining.
The SQL/MM API is a standardized open SQL API. It supports the
following items:
- Exploits object-relational extensions of SQL
- Objects are implemented as user-distinct types and user-defined
structured types
- These objects are stored in large object columns (LOB) in the
relational tables
- Methods invoked on those objects are implemented as user-defined
methods (UDM) and user-defined functions (UDF)
- Enables data mining users to use the SQL language and DB2 optimizer
- Runs functions directly on the database server, where the data
is stored
The following list of data types shows the SQL API terminology:
- MiningData
- Describes physical schema of input table
- LogicalDataSpec
- Describes the logical schema of input table
- RuleFilter
- Describes the rule filter constraints that limit the rule length
and the rule content
- Settings
- Describes the function parameters for the modeling phase
- BldTask
- Describes the job parameters for the modeling phase
- TestTask
- Describes the job parameters for the test phase
- Model
- Describes the rules and patterns found in the modeling phase
- TestResult
- Describes the quality information found in the test phase
Note: The database objects are created in the predefined schema IDMMX,
which is provided by Intelligent
Miner.