Baidu plans to announce a sweeping expansion of its partnership with graphics chip maker Nvidia on Wednesday, bringing Nvidia’s latest chips into Baidu’s cloud services for artificial intelligence research and optimizing Baidu’s deep learning system.
The two companies have been working together on AI-related projects for a few years now, and the new agreement, scheduled to be announced during the Baidu Create 2017 AI developer conference, expands on much of that existing work.
A core MBA class at Missouri University of Science and Technology prepares students for this distinct possibility, and teaches them how to coexist with their future artificial intelligence colleagues.
Dr. Keng Siau introduced artificial intelligence and machine learning into his business curriculum during the spring 2017 semester.
The Artificial Intelligence, Robotics, and Information Systems Management course looks at the latest developments in artificial intelligence, machine learning, robotics, automation and advanced information technology, and “their effect on our current ways of life and work as well as on economic/business models,” says Siau, professor and chair of the business and information technology department. The course will be offered again in spring 2018.
Researchers at Carnegie Mellon University have developed a computer that can understand and track body movements in real time, including individual fingers.
The computer could allow robots to read body language, helping them better perceive your mood and interact with you in social settings, said Yaser Sheikh, an associate professor of robotics leading the project. Robots could better understand what we mean when we point at something or do other hand gestures like putting our finger to lips to tell someone to keep quiet.
Big dreams need big data. And former Vice President Joe Biden believes his National Cancer Moonshot initiative needs lots of it — shared in ways like never before — to find a cure.
The initiative aims to accelerate research efforts and break down barriers to progress by enhancing data access and facilitating collaborations with researchers, doctors, philanthropies, patients and patient advocates, and biotechnology and pharmaceutical companies.
Groups started as a way of sharing things with smaller circles of people in your life—your family, your book club, your kickball team—without cluttering your profile. Nowadays, though, you can find a group for almost anything. Zuckerberg claims there are 100 million people in “meaningful” groups that they consider vital to their daily lives. He wants to see that number hit 1 billion within five years.
To get there, Facebook plans to deploy—what else?—artificial intelligence to recommend groups you may find “meaningful.” More importantly, the company added tools to schedule posts, screen and block members, and link groups. Together, these additions will help you find groups you might want to join, and to give the people leading those Groups greater control over their content and membership.
In a revolution that extends across much of science, researchers are unleashing artificial intelligence (AI), often in the form of artificial neural networks, on the data torrents. Unlike earlier attempts at AI, such “deep learning” systems don’t need to be programmed with a human expert’s knowledge. Instead, they learn on their own, often from large training data sets, until they can see patterns and spot anomalies in data sets that are far larger and messier than human beings can cope with (see boxes, pp. 20, 23, 25, 26, & 27). [Introduction for Science special issue on Artificial Intelligence]
The CEO of Baidu, Robin Li, arrived at his company’s first AI developer conference, held in Beijing this week, in a vehicle that has the potential to reshape the world of self-driving cars.
The vehicle was controlled using software that Baidu (50 Smartest Companies 2017) plans to offer for free in the coming years through a project called Apollo. By making the brains of a self-driving car available to anyone, the Apollo project could help China’s many young carmakers get up to speed rapidly.
It also reflects China’s broader ambition to establish itself as a leading hub of artificial intelligence.
App-based ride services have changed the urban world, often for the good. San Francisco already has some 45,000 Uber and Lyft drivers cruising the city’s busiest areas. In New York City, as of last fall, ride-service companies carried 15 million passengers a month — a tripling of ridership in a year and a half.
And yet many city, county and regional planners have precious few ways to measure the effects of this transportation revolution. They simply do not know the number of cars out there, the mileage they log, where they’re driving people, and how much people are paying for the ride.
Soon after President Donald Trump took office with a pledge to cut regulations, Republicans in Congress killed an Obama-era rule restricting how broadband companies may use customer data such as web browsing histories.
But the rule may be finding new life in the states.
Lawmakers in almost two dozen state capitols are considering ways to bolster consumer privacy protections rolled back with Trump’s signature in April. The proposals being debated from New York to California would limit how AT&T Inc., Verizon Communications Inc. and Comcast Corp. use subscribers’ data.
We’re excited to announce we closed on $30 million in funding from our world-class team of investors, including Accel, Jeff and Tricia Raikes, and Nelnet.
We plan to use the funds to pair innovations in machine learning and computer vision with our in-house group of professional analysts, providing teams around the world quicker access to the insights they need for training and game preparation.
The Regenstrief Institute, internationally recognized for its research and training programs in clinical informatics, will now train researchers in the increasingly important fields of public and population health informatics. The unique new program, in collaboration with Indiana University School of Medicine and IU’s Richard M. Fairbanks School of Public Health at Indiana University-Purdue University Indianapolis, is supported by a five-year, $2.5 million award from the National Library of Medicine, an institute of the National Institutes of Health.
The Indiana Training Program in Public and Population Health Informatics, commencing in July 2017, will prepare graduate students and post-doctoral fellows to work in a broad spectrum of entities in the healthcare industry and academia, as well as for local, state and federal public health departments. These trainees will fill a need — forecasted to grow over the next decade and beyond — for informaticians who can design, validate and implement solutions key to the maintenance and improvement of human health.
In an interview for Flux, I sat down with today’s modern Josephine Cochrane, Helen Greiner, the co-founder of iRobot.
The company behind the first automated and commercially successful home vacuum, the Roomba, iRobot’s appliance hit the market in 2002 and has now sold more than 16 million units worldwide.
We got into how founders should think about timing a market, navigating user adoption cycles and iterating on product. In 2008, Helen made the jump from terrestrial to aerial robotics, founding drone company CyPhy Works to focus on applications including public safety, construction and agriculture. Helen also shared her thoughts on why the sky is a natural superhighway for drone delivery, how to get more women into technology, her love of Star Wars and what happened when she tried to fly her drone on the White House lawn.
An excerpt of our conversation is published below.
A deep-learning algorithm developed by EPFL scientists can generate melodies that imitate a given style of music. The “deep artificial composer” could one day generate convincing music for multiple instruments in real time, with applications ranging from video games to helping composers in the creative process.
La Jolla, CA October 5-6. A thoughtful exploration of the clinical evidence necessary to drive the widespread uptake of digital health solutions will be the focus of Scripps Translational Science Institute’s 2017 Digital Medicine conference. [$$$]
To apply, please submit a 3-page proposal that describes your goals, expected outcomes, a tentative timeline, collaborators, and a very high level budget at ropensci.org/fellowships. As part of your submission form, you will also have to provide details about your affiliation, PI status, and the name of your PI (if you cannot serve in that role). Deadline for applications is September 1.
The Adam optimization algorithm is an extension to stochastic gradient descent that has recently seen broader adoption for deep learning applications in computer vision and natural language processing.
GCP offers a range of compute services that go from giving users full control (i.e., Compute Engine) to highly-abstracted (i.e., Firebase and Cloud Functions), letting Google take care of more and more of the management and operations along the way. Here’s how many long-time readers of our blog think about GCP compute options.
“We’ve designed a new convolutional neural network structure for mobile platforms which utilizes pointwise group convolution and channel shuffle. Under the budget of 40MFLOPS, we’ve achieved 6.7% absolute top-1 error reduction on ImageNet classification compared to MobileNets. Empirically, our network with approximately the same error runs 13x faster than AlexNet on an ARM platform.”
Feature extraction is the process of breaking down a user’s real world (as seen by on-device sensors) into small, informative chunks, or features. These features are then used as training and test input to downstream neural nets, which make predictions. Traditionally, this is done “offline” via post-processing of a (usually very large) dataset collected and stored on a server somewhere. However, since our mission is to keep user data on the device, we can’t use this server-side approach. Instead, we need a way to efficiently query streaming data on-device, without having knowledge of the entire stream.
As it turns out, a class of algorithms have already been developed for this type of problem: Data Sketches. In this post, we’ll offer a brief introduction to data sketching in general, then talk a bit about Set’s specific approach to sketching location data in particular, in later posts.
“At MondoDB World 2017 recently, Jay Runkel, principal solutions architect at MongoDB, demonstrated how to apply a little math to get a pretty close guesstimate of the resources needed to run your database workload.”
In general, individual epoch observations and transits will be released only with the final catalogue. However, when variable star solutions, etc. are released, relevant subsets of epoch data may also be released. Additionally, where the release of epoch data is of immediate scientific interest, such release will be made on an individual source-by-source basis through the science alerts.