IBM to Apply Machine Learning Algorithms to Local Data

    One of the longest standing tenets of computing is that it’s always better to bring the code to the data than the other way around. In recognition of that fundamental principle, IBM today announced that it will start making it possible for IT organizations to apply the machine learning algorithms it has developed to data wherever it resides.

    Thus far, most of IBM’s machine learning efforts have been focused on building a suite of IBM Watson services. Now IBM is pledging to give customers the ability to apply machine learning algorithms to data without requiring them to move data from their local data center into the cloud.

    Dinesh Nirmal, vice president of analytics development for IBM, says IBM will initially make these algorithms available on mainframes running its z/OS operating system.

    “The mainframe is still the place where the most amount of data is stored in the enterprise,” says Nirmal.

    Following the delivery of that capability, IBM will then focus on all the other platforms it supports. To facilitate those efforts, IBM has created Cognitive Automation for Data Science from IBM Research to help organizations automate the process that data scientists use to identify the right algorithm by scoring their data to identify the best available algorithm to match their needs, says Nirmal.

    The goal, says Nirmal, is to enable IT organizations to create their own operational analytic models using any programming language or machine learning framework.

    In general, Nirmal says, IBM Watson services are best used in the development of applications based on what are known as deep learning algorithms that have a foundation in neural networking technology. Other classes of machine learning algorithms are better suited for being deployed against transactional data in real time. IBM expects that, over time, organizations will make use of both kinds of algorithms to create a new generation of intelligent applications.

    The difference now is that, instead of having to go to the trouble of moving data into the cloud to employ those algorithms, an IT organization will soon be able to apply them against new and existing legacy applications to infuse advanced analytics throughout the enterprise.

    Mike Vizard
    Mike Vizard
    Michael Vizard is a seasoned IT journalist, with nearly 30 years of experience writing and editing about enterprise IT issues. He is a contributor to publications including Programmableweb, IT Business Edge, CIOinsight and UBM Tech. He formerly was editorial director for Ziff-Davis Enterprise, where he launched the company’s custom content division, and has also served as editor in chief for CRN and InfoWorld. He also has held editorial positions at PC Week, Computerworld and Digital Review.

    Latest Articles