IBM SPSS Modeler and Database Integration modeling and optimization (i.)

IBM SPSS Modeler and database integration and configuration

IBM SPSS Modeler is a set of data mining tools that, as an important part of the IBM Analysis and prediction solution, can use commercial technology to quickly build predictive models and apply them to business activities to improve the decision-making process. It can process and model enterprise-class massive data, through powerful database integration function, can directly with the enterprise existing database integration data mining. Not only to avoid the enterprise capital duplication, but also to obtain better data mining performance.

For example, a company after years of accumulation, there are very large data and stored in the database, I hope to use SPSS Modeler in the existing data mining to make the decision in favor of the company. Then the company will be faced with some questions or questions, including: How SPSS Modeler communicate with the database, how to obtain data for modeling, how to store modeling results, how to ensure the performance of operations on large data, and so on.

This series of articles will be divided into three parts to answer each of these questions, the first part of the introduction of basic knowledge including database configuration and operations, the second part of the database integration modeling, the third part of the performance optimization. This is the first part.

Installing drivers

SPSS Modeler can import data from multiple databases using ODBC (Open database connections) through the database source node, including dozens of databases such as DB2, Netezza, Oracle, Teradata, Microsoft SQL Server, and so on. To read or write to the database, you must install the driver package for the related database, configure the ODBC data source, and configure read or write permissions as needed. The IBM SPSS Data Access Pack contains a set of ODBC drivers for this purpose and supports a variety of operating system platforms.

IBM SPSS Modeler for the typical C/s architecture products, if only in the local (stand-alone) mode to run IBM SPSS Modeler, you must install the driver on the local computer.

If you are connecting to a remote IBM SPSS Modeler Server in distributed mode to run SPSS Modeler, you need to install the ODBC driver on the computer that installs the SPSS Modeler server

Use the following general steps to access data in the database:

Install the ODBC driver for the database that you want to use and configure the data source.

In the Database Node dialog box, connect to the database using either table mode or SQL query mode.

Select the table from the database.

Using the tabs in the Database Node dialog box, you can change the usage type and filter the data fields.

These steps are described in more detail in later chapters. This is the first driver installation and configuration.

Windows platform Database driver installation and data source configuration

The Windows version of the IBM SPSS Data Access Pack is available in 32-bit and 64-bit versions, and we use 32-bit demos here, and please be careful to choose the right version of the installation when you actually use it.

Its installation process takes a typical step-by-step method, we only need to use the default settings step-by-step installation. After the installation is complete, open the Control Panel-> management tool-> Data Source (ODBC), and you can see that a batch of corresponding database drivers are installed on the driver page.

Figure 1.ODBC Driver

We use DB2 as an example to continue with the following operations, similar to other databases.

Back to the ODBC database Source Manager System DSN page, click the Add button to select the SPSS Inc OEM 6.0 DB2 Wire Protocol driver.

Figure 2.ODBC Data Source Manager-System DSN