IBM launches machine learning analytics product for private cloud
Thu 16 Feb 2017
IBM has announced the launch of IBM Machine Learning, a cognitive platform for the z System mainframe. The new platform uses the core technology from IBM Watson to perform big data analysis for clients using z System for cloud data storage.
The new Machine Learning platform automates the creation and deployment of analytic models, allowing data scientists to perform analysis in-house. Models can be written in any language (Scala, Java, Python), using any framework and data type required for best results.
IBM Machine Learning also includes a function called Cognitive Automation for Data Scientists, which helps data scientists to choose the correct algorithm for their needs. The Cognitive Automation function provides a best match from a list of available algorithms, taking into account the purpose of the model and speed constraints.
“Machine Learning and deep learning represent new frontiers in analytics. These technologies will be foundational to automating insight at the scale of the world’s critical systems and cloud services,” said Rob Thomas, General Manager, IBM Analytics.
While Machine Learning will initially be available on the z System for private cloud data, the company believes that the service will soon expand to hybrid and public cloud implementations, accelerating the adoption both of hybrid and public cloud solutions and of machine learning analytics worldwide.
IBM envisions applications for the Machine Learning platform in retail, financial services, and healthcare industries. IBM Machine Learning was evaluated in a pilot program by Argus Health, a pharmacy and health care solutions provider. Argus used the platform in concert with its existing pharmacy claims platform to create analytic models to help manage pharmacy costs.
The IBM z System mainframe is currently in use by 92 of the world’s largest 100 banks. Applications of machine learning in the financial sector could include up-to-date investment recommendations that take into account current trends without a lag time for processing analytics.
The z System Mainframe is also used by 70% of the world’s largest retailers, and 23 of the top 25 airlines globally. The mainframe is capable of processing up to 2.5 billion transactions per day, and the implementation of IBM Machine Learning means that data analysts can harness that processing power without having to move data off the mainframe for analysis.