Data Mining Extensions
Like SQL, it supports a data definition language, data manipulation language and a data query language, all three with SQL-like syntax. Whereas SQL statements operate on relational tables, DMX statements operate on data mining models. Similarly, SQL Server supports the MDX language for OLAP databases. DMX is used to create and train data mining models, and to browse, manage, and predict against them. DMX is composed of data definition language (DDL) statements, data manipulation language (DML) statements, and functions and operators.
DMX Queries are formulated using the
They can extract information from existing data mining models in various ways.
Data definition language
The data definition language (DDL) part of DMX can be used to
- Create new data mining models and mining structures -
CREATE MINING STRUCTURE, CREATE MINING MODEL
- Delete existing data mining models and mining structures -
DROP MINING STRUCTURE, DROP MINING MODEL
- Export and import mining structures -
- Copy data from one mining model to another -
Data manipulation language
The data manipulation language (DML) part of DMX can be used to
- Train mining models -
- Browse data in mining models -
- Make predictions using mining model -
SELECT ... FROM PREDICTION JOIN
Example: a prediction query
This example is a singleton prediction query, which predicts for the given customer whether she will be interested in home loan products.
SELECT [Loan Seeker], PredictProbability([Loan Seeker]) FROM [Decision Tree] NATURAL PREDICTION JOIN (SELECT 35 AS [Age], 'Y' AS [House Owner], 'M' AS [Marital Status], 'F' AS [Gender], 2 AS [Number Cars Owned], 2 AS [Total Children], 18 AS [Total Years of Education] )