IBM PureSystems Centre

Extending the value of IBM PureSystems

Universal PMML Plug-in for Netezza

IBM PureData System
Provided by:Zementis
Request Information

The Zementis Universal PMML Plug-in for IBM PureData® System for Analytics delivers standards-based execution of predictive analytics for in-database scoring. It allows for predictive solutions built in any PMML-compliant data mining environment to be instantly deployed and executed close to your data, inside the IBM PureData appliance.
  • - Business Value

    Pattern Overview

    Zementis and IBM PureData® System for Analytics help companies easily deploy, execute and integrate scalable standards-based predictive analytics. This joint solution combines the Zementis Universal PMML Plug-in (UPPI) for in-database execution of predictive analytics with IBM PureData System for Analytics (powered by Netezza® technology). UPPI for IBM PureData System for Analytics fuses in-database scoring and data warehousing into a scalable, high-performance, massively parallel analytic platform that easily processes through petascale data volumes. UPPI takes full advantage of the high-performance data warehouse asymmetric with its massively parallel processing (AMPP) architecture for rapid execution of standards-based predictive analytics.

    Universal PMML Plug-in: Features

    The Universal PMML Plug-in fully supports the Predictive Model Markup Language (PMML), the de-facto standard for data mining applications. PMML, developed by the Data Mining Group, provides a standard way for an application to define predictive analytic solutions so that they can be easily shared with any other application that supports PMML.

    Business Case

    Business Problem

  • -Technical Details

    With PMML, the Plug-in delivers a wide range of predictive analytics for high performance scoring, including: • Decision Trees for classification and regression • Neural Network Models: Back-Propagation, Radial-Basis Function, and Neural-Gas • Support Vector Machines for regression, binary and multi-class classification • Linear and Logistic Regression (binary and multinomial) • Naïve Bayes Classifiers • General and Generalized Linear Models • Cox Regression Models • Rule Set Models (flat decision trees) • Clustering Models: Distribution-Based, Center-Based, and 2-Step Clustering • Scorecards (including reason codes) • Association Rules • Multiple Models: Model ensemble, segmentation, chaining and composition

  • - More Information

    Provider Information

    Name: Henry Huang


    Phone: +1 (619) 330 0780 x 2000

PureSystems patterns are part of a broad portfolio of PureSystems solutions for business, cloud, and infrastructure applications.

See more solutions

Related tagged content



You must  login with your IBM ID to be able to bookmark. Select "Login" to bookmark or "Close" to return.