The
first steps in a mining project are to consolidate
the data to be analyzed into a data mart and to transform it into
the required format for the mining algorithms.
With
the Data Source Explorer in the Design Studio, you can perform
the following tasks:
- Creating database connections to JDBC
data sources, for example, Db2® databases
- Navigating the schemas, tables, or views to understand the database
structure
- Displaying sample contents of database tables
- Exploring
value distributions or field relationships in tables
- Exploring
and visualizing data mining models that are stored in
the database
- Exporting data mining models as PMML files
- Importing
data mining models from PMML files
- Copying data mining models
from one database to another
- Enabling
databases for mining and thus creating
the stored procedures and user-defined functions for Intelligent Miner®
With the data design features, you can create
new tables for your
mining data mart. For example, you might want to perform the following
tasks:
- Creating a physical database design
- Forward
engineering the design into a connected database
- Reverse engineering
and modifying a database design from an existing
database
With the SQL warehousing and
data mining features,
you can create data flows and mining flows to
perform the following tasks:
- Importing data from Db2-
or non-Db2 databases by using JDBC connections
- Transforming
and preprocessing data by using SQL-based transform
operators
- Preprocessing data by using the Data Preparation
feature