Google Open Sources Machine Learning Library TensorFlow - InformationWeek

InformationWeek is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Data Management // Big Data Analytics
05:25 PM
Connect Directly

Google Open Sources Machine Learning Library TensorFlow

With TensorFlow now open sourced by Google, companies and the research community can implement machine learning systems more easily and more efficiently.

Cognitive Computing Powers 6 Smart Deployments
Cognitive Computing Powers 6 Smart Deployments
(Click image for larger view and slideshow.)

Google may be focused on machine learning, but its researchers understand that human learning is a prerequisite for smarter machines.

To help people inside and outside the company learn how to teach machines more efficiently, Google on Monday released its latest machine learning library, TensorFlow, under the open source Apache 2.0 license. The software can turn machine learning algorithms that have been written as graphs of symbolic expressions into efficient low-level code.

"TensorFlow is great for research, but it's ready for use in real products too," explained Jeff Dean, senior Google Fellow, and Rajat Monga, technical lead, in a blog post. "TensorFlow was built from the ground up to be fast, portable, and ready for production service. You can move your idea seamlessly from training on your desktop GPU to running on your mobile phone."

Dean and Monga noted that TensorFlow is Google's second-generation machine learning system, following in the footsteps of DistBelief, a deep learning system developed in 2011. The company used DistBelief to teach its computers how to recognize cats in unlabeled YouTube images, to improve speech recognition accuracy by 25%, and to build the image search in Google Photos. And, it's using TensorFlow in its Smart Reply service for Inbox by Gmail.

(Image: Google)

(Image: Google)

The most significant difference between the two systems is that TensorFlow isn't tied to Google's IT infrastructure. Absent any requirement for a specific hardware configuration, TensorFlow can be employed by anyone with the appropriate technical background and modest IT assets.

Machine learning is central to many of the most popular cloud computing services today and it's likely to become even more significant in the years to come. As Google's researchers explain in a paper describing TensorFlow, machine learning is useful in more than a dozen areas of computer science and other disciplines, including speech recognition, computer vision, robotics, information retrieval, natural language processing, geographic information extraction, and computational drug discovery.

The potential utility of TensorFlow may not be obvious to those outside of computer science, but the fact that Google CEO Sundai Pichai announced the release on the Google Official Blog indicates that this isn't esoteric research. It's near and dear to Google's engineers.

On Google's Q3 2015 earnings call, held under the name of its parent company Alphabet, Pichai referred several times to the significance of machine learning. "Machine learning is a core, transformative way by which we're rethinking everything we're doing," he said, noting that he expects to apply machine learning systematically to search, ads, YouTube, and other products.

[Read Machine Learning: How Algorithms Get You Clicking.]

In an email, Patrick Ehlen, chief scientist at Loop AI Labs, said that machine learning allows companies to use unstructured data to reveal things about their business that went unappreciated.

"By creating algorithms, these models allow computers to find insights that were previously difficult or impossible for humans to find," said Ehlen. "Furthermore, we see the future of this technology in unsupervised learning, where the system is not told the 'right answer' for any of the data, and the algorithms figure out the underlying structure of the data all on its own."

Thomas Claburn has been writing about business and technology since 1996, for publications such as New Architect, PC Computing, InformationWeek, Salon, Wired, and Ziff Davis Smart Business. Before that, he worked in film and television, having earned a not particularly useful ... View Full Bio

We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.
Comment  | 
Print  | 
More Insights
InformationWeek Is Getting an Upgrade!

Find out more about our plans to improve the look, functionality, and performance of the InformationWeek site in the coming months.

Becoming a Self-Taught Cybersecurity Pro
Jessica Davis, Senior Editor, Enterprise Apps,  6/9/2021
Ancestry's DevOps Strategy to Control Its CI/CD Pipeline
Joao-Pierre S. Ruth, Senior Writer,  6/4/2021
IT Leadership: 10 Ways to Unleash Enterprise Innovation
Lisa Morgan, Freelance Writer,  6/8/2021
White Papers
Register for InformationWeek Newsletters
Current Issue
Planning Your Digital Transformation Roadmap
Download this report to learn about the latest technologies and best practices or ensuring a successful transition from outdated business transformation tactics.
Flash Poll