Mainframe AI (IT Toolbox Blogs)

Those of you who enjoy scifi movies will know that as soon you install self-learning software on a big enough computer, it will become self-aware and takes over the world. You’ll remember from the 1991 Terminator film that “Human decisions are removed from strategic defence. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14am. Eastern time, August 29th.”

Meanwhile, back in the real world, we find that IBM is adding the Machine Learning technology from Watson to its z/OS mainframe for smarter, faster analytics of transaction data. The idea is that it now becomes possible to analyse mainframe data in place. And, although it uses expensive CPU cycles, the analysis is faster than any of the alternatives (such as extracting and running on Hadoop under Linux).

According to IBM: “IBM Machine Learning for z/OS … helps organizations quickly ingest and transform data to create, deploy, and manage high-quality self-learning behavioural models using IBM z Systems data, securely in place and in real time.”

Originally called IBM Predictive Analytics, and based on SPSS, IBM launched Machine Learning late last year on its BlueMix cloud platform. IBM Machine Learning helps automate the creation, training, and deployment of operational analytic models that will support:

  • Any language (eg Scala, Java, Python),
  • Any popular Machine Learning framework like (eg Apache SparkML, TensorFlow, H2O),
  • Any transactional data type.

IBM tells us that this will simplify the work of data scientists in analytic model creation, deployment, and management and help ensure model accuracy because it “continuously analyses the data and models to provide better predictions and optimization of behavioural models, speeding time to insights”.

It makes use of the IBM Research-produced Cognitive Assist for Data Science. This, IBM confirms, will “assist data scientists in choosing the right algorithm for the data by scoring their data against the available algorithms and providing the best match for their needs. The service also considers various circumstances, such as what the algorithm is needed to do and how fast it needs to produce results.”

It’s not just unstructured data from Big Data that can be analysed, it can also include structured data from Data Warehousing and Business Intelligence products. Previously, these would have needed to go through the process of extraction, transformation, and loading before they can be used. And, because of this CPU-intensive process, valuable mainframe data is often left unused. The introduction of this Watson technology will makes that information more easily available.

IBM suggests that there will be numerous applications that it’s Machine Learning system can be used for, such as in retail sales it can compute daily market trends, in finance it could create up to the hour details for advisers and brokers, and in healthcare it could track personalized offerings for individual patients connected with IoT devices.

Prior to this announcement, organizations would need to move their data to IBM’s BlueMix cloud and lease computing services if they wanted to use this facility. Now, they’ll be able to construct and deploy these applications on their own systems, which means they’ll have to pay for their own CPU cycles, but it will remove the cost, latency, and risk of moving data off premise that comes with a cloud deployment. Organizations will, effectively, have their own private cloud.

General Manager, IBM Analytics, Rob Thomas said: “Over 90 percent of the data in the world can’t be Googled. It resides behind firewalls on private clouds. How do we automate intelligence?

“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.

“IBM Machine Learning was designed leveraging our core Watson technologies to accelerate the adoption of machine learning where the majority of corporate data resides. As clients see business returns on private cloud, they will expand for hybrid and public cloud implementations.”

A couple of other things worth noting are that IBM is particularly enthusiastic about the importance of private clouds to industries that manage a lot of sensitive data, like healthcare. Secondly, after z/OS, IBM Machine Learning will be available on POWER servers and then other platforms – which may mean x86 servers.

So, perhaps not becoming self-aware, yet.

Find out more about iTech-Ed here.

Predictive Analytics

Source: SANS ISC SecNewsFeed @ February 26, 2017 at 04:12AM