• Skip to primary navigation
  • Skip to main content
  • Skip to primary sidebar
  • Skip to footer

My TechDecisions

  • COVID-19 Update
  • Best of Tech Decisions
  • Topics
    • Video
    • Audio
    • Mobility
    • Unified Communications
    • IT Infrastructure
    • Network Security
    • Physical Security
    • Facility
    • Compliance
  • RFP Resources
  • Downloads
  • Podcasts
  • Subscribe
  • Project of the Week
  • About Us
    SEARCH
IT Infrastructure, News

Making the NAS Algorithm More Accessible: This MIT Research Could Be a Boon to Machine Learning

Neural networks are a useful machine learning system, but they take up a lot of GPU power...At least, usually. This MIT research could change that.

May 31, 2019 TechDecisions Staff Leave a Comment

neural networks, MIT research

One of the latest trends in machine learning revolves around using algorithms to automatically create neural networks, a type of machine learning system, according to a post from MIT news. Doing so could be more accurate than humans, but it is very “computationally expensive.” So why do it?

One of these “neural architecture search (NAS)” algorithms developed by Google took 48,000 GPU hours to produce one neural network, the MIT story says. That network was then used for tasks related to image classification. And while Google has the means to do that, the necessary hardware is beyond reach for most.

A NAS Algorithm That Can ‘Beat’ Google’s

The MIT article references a soon-to-be-presented research paper which theorizes on an algorithm that can directly learn neural networks in only 200 GPU hours, rather than 48,000.

This could enable smaller developers to more easily design neural network architectures which run fast on hardware applications, the article says.

“In their work, the researchers developed ways to delete unnecessary neural network design components, to cut computing times and use only a fraction of hardware memory to run a NAS algorithm. An additional innovation ensures each outputted CNN runs more efficiently on specific hardware platforms — CPUs, GPUs, and mobile devices — than those designed by traditional approaches. In tests, the researchers’ CNNs were 1.8 times faster measured on a mobile phone than traditional gold-standard models with similar accuracy.” — Post from MIT news

Related: Google AI’s Translatotron Can Translate A Speaker’s Voice — With the Same Characteristics

For each network layer, the NAS algorithm uses information sampled from the architecture to create another architecture that runs as quickly as possible, the article says. In those experiments, the researchers found it ran twice as fast as a gold-standard model on mobile devices.

Tagged With: Machine Learning

Related Content:

  • Avocor Aquarius WorkSpace Intelligence, Meeting Room Data Analytics platform Avocor WorkSpace Intelligence Meeting Room Data Analytics Service
  • Malwarebytes SolarWinds Malwarebytes CEO Says SolarWinds Attackers Accessed Internal Company…
  • Biden IT Biden Administration Calls For IT Experts In White…
  • Biden IT Funding President Biden Wants To Invest $10B in U.S.…

Free downloadable guide you may like:

  • Top 9 Reasons Enterprise IT Leaders Are Moving Their Video Surveillance to the Eagle Eye Cloud

    Working in IT has enough challenges without adding in the complications of surveillance video. Things like total cost of maintenance, how the VMA manages bandwidth, what cameras are supported, what level of cybersecurity is provided, and what integrations are available to use are important factors IT managers have to think about when assessing a video […]

Reader Interactions

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Get the FREE Tech Decisions eNewsletter

Sign up Today!

Latest Downloads

9 Technology Products to Help Combat COVID-19 Spread in the Workplace

As the Coronavirus continues on and leads us further into uncertainty, the question remains, “when do we return to the office?” For some the answer...

Top 9 Reasons Enterprise IT Leaders Are Moving Their Video Surveillance to the Eagle Eye Cloud

Working in IT has enough challenges without adding in the complications of surveillance video. Things like total cost of maintenance, how the VMA m...

Using Live Chats and Chatbots to Increase Customer Engagement

There's a lot to consider when building out a chatbot experience to ensure that it delivers a seamless experience and meet your business goals.

View All Downloads

Would you like your latest project featured on TechDecisions as Project of the Week?

Apply Today!
Sharp Microsoft Collaboration HQ Logo

Learn More About the
Windows Collaboration Display

More from Our Sister Publications

Get the latest news about AV integrators and Security installers from our sister publications:

Commercial IntegratorSecurity Sales

Footer

TechDecisions

  • Home
  • Welcome to TechDecisions
  • Subscribe to the Newsletter
  • Contact Us
  • Media Solutions & Advertising
  • Comment Guidelines
  • RSS Feeds
  • Terms of Use
  • Privacy Policy
  • Twitter
  • Facebook
  • Linkedin

Free Technology Guides

FREE Downloadable resources from TechDecisions provide timely insight into the issues that IT, A/V, and Security end-users, managers, and decision makers are facing in commercial, corporate, education, institutional, and other vertical markets

View all Guides
TD Project of the Week

Get your latest project featured on TechDecisions Project of the Week. Submit your work once and it will be eligible for all upcoming weeks.

Enter Today!

© 2021 Emerald X, LLC. All rights reserved.