Machine learning, which provides a system with the capability of learning things without being prompted and pushed by external programs, is gaining tremendous popularity these days. And now, global tech and innovation company IBM is planning to present machine learning for its mainframe users. This AI will later be incorporated with all other technologies having their bulk of data hidden behind a “private cloud” or the company’s firewall.
In case you were surprised to hear the name of mainframe computers, which went out of popular fashion years ago, these computers are still being widely used in several world renowned banks and industries, airlines and insurance firms. IBM says that its systems mainframe can easily process roughly 2.5 billion transactions a day.
On a bigger frame, IBM is actually trying to bring the brains well-versed with Watson machine learning to its mainframe customers, and later to all the techs in the data center, so as to gain access to more of machine learning data and then take its advantage.
IBM analytics general manager Rob Thomas says;
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 [for these data sources]?
Data scientists find excellent machine learning capabilities in the cloud and IBM is trying to provide them with the same features over mainframe. The purpose is to dehumanize the rather monotonous job of creating, testing and deploying analytical models. The solutions are compatible with regular open source languages like Scala, Java and Python, and machine learning frameworks like Apache SparkML, TensorFlow and H2O.
Along with the open source tools, IBM is also offering Cognitive Assist for Data Science from IBM Research, which will help you choose the best algorithm from all the available ones by comparing them to check which one best fits the users requirements.
This allows data scientists to build a model and IBM Machine Learning technology will choose the best algorithm. It then builds a feedback loop because as more data comes in, the algorithm gets updated and gets smarter.
Though in a crude state, present days artificial intelligence and machine learning have been a part of mainframe since decades. Thomas tells that companies will be able to obtain economical advantage through these tools as well. He also stressed that processing data on the mainframe with the help of provided tools will be more practical and economical in comparison to the processing of same data on the cloud.
The enhanced capabilities will roll out later this quarter for mainframe users.
