CHICAGO, Nov. 27, 2019 /PRNewswire-PRWeb/ — The Data Mining Group, a vendor-led consortium of companies and organizations developing standards for statistical, data mining and machine learning models, announced today the release of version 4.4 of the Predictive Model Markup Language (PMML). PMML is an application and system independent XML interchange format for statistical, data mining and machine learning models. The goal of the PMML standard is to encapsulate a model independent of applications or systems in such a way that two different applications (the PMML Producer and Consumer) can use it.

“As Artificial Intelligence (AI) becomes a core component of IT and IoT solutions, the ability to streamline the implementation of machine learning models – from development to operational deployment – through a common industry standard ensures the scalability of data science.” says Johannes Viegener, SVP R&D IoT Analytics at Software AG. “By incorporating Time Series and Anomaly Detection with its latest release, PMML amplifies its benefits for the Internet of Things (IoT) and continues to empower organizations to deploy, audit, integrate and execute predictive models with greater ease and at a lower cost.”

Some of the elements that are new to PMML v4.4 include:

• Addition of time series model subtypes for ARIMA, State Space models, GARCH
• Anomaly Detection model
• Improved documentation for several models
• New mathematical built-in functions
• Improved support for invalid and missing values

Todd Moore, IBM VP of Open Technology said “I am excited to see the release of PMML 4.4, as IBM is one of the founders and most active members of the Data Mining Group. PMML support is integral to both traditional IBM offerings, such as IBM SPSS Statistics and IBM SPSS Modeler, and newer, such as Watson Studio (Cloud, Local, and Desktop versions) and Watson Machine Learning. Used as both an internal model and transformation representation, and to import/export models between our products and third party platforms, including open source, release of PMML 4.4 adds significant new features. These include ARIMA, State Space models, GARCH for Time Series model, Anomaly Detection, and greatly improves the documentation for Bayesian Network scoring, and other areas.”

“With the growing importance of deep learning and AI, the ability to share machine learning models across applications is critical. PMML v4.4 provides this capability for AI platforms, systems and frameworks,” says Robert L. Grossman, a Data Mining Group Director.

About PMML:

PMML is the leading standard for statistical, data mining and machine learning models and is supported by multiple prominent vendors and organizations. With PMML, it is straightforward to develop a model on one system using one application and deploy the model on another system using another application.

About DMG:

The Data Mining Group (DMG) is an independent, vendor led consortium that develops analytic standards, such as the Predictive Model Markup Language (PMML) and the Predictive Format for Analytics (PFA).

For more information about the Data Mining Group and the PMML standard, go to: http://www.dmg.org

SOURCE Data Mining Group