If election security is an engineering problem, the Defense Advanced Research Projects Agency is heading to the right place to solve it. The Pentagon’s blue skies projects agency is taking its System Security Integrated Through Hardware and Firmware (SSITH) to the 2019 DEF CON hacking conference to demonstrate its capabilities before the dark lords and apprentices of the underground community.
SSITH will be on display as part of the conference’s Voting Village, where researchers will explore what can and cannot be done to interfere with voting machines and, by extension, elections.
“We expect the voting booth demonstrator to provide tools, concepts and ideas that the election enterprise can use to increase security; however, our true aim is to improve security for all electronic systems. This includes election equipment, but also defense systems, commercial devices and beyond,” said Dr. Linton Salmon, the program manager leading SSITH, in a release from DARPA.
Last week was a brutal one for 2U, the online program management company that had been the darling of investors for much of the last decade. As its officials acknowledged for the second time in three months that the company would grow more slowly than it had previously predicted, stockholders bailed in stunning numbers, reducing 2U’s value by roughly two-thirds (and $1.5 billion) in 24 hours.
Today it’s back to business as usual, starting what is likely to be a long process of reassuring investors (and perhaps the entire higher education technology landscape) that 2U can and will continue to be a vibrant, creative leader in the space.
The company announced this morning that it will start its first-ever undergraduate degree program, an online bachelor of science in data science and business analytics, in conjunction with the London School of Economics and Political Science and its parent, the University of London.
The ability to acquire language is often taken for granted. From the utterance of a first word, to the eventual sentences of early childhood, language quickly comes to encode our thoughts and enable our interpersonal communications. Language is so fundamental, it seems almost instinctual. Yet, for computers, absorbing natural language is incredibly difficult. Where a child could easily interpret the meaning of a novel phrase, “to run slowly,” knowing both the definitions of “to run” and “slowly,” a computer struggles to make these same, seemingly intuitive, connections. It is important to understand how the mind handles compositionality in language and thought in order to attempt to reproduce these abilities in machines.
In their recent publication, Brenden M. Lake, Assistant Professor of Psychology and Data Science, and Facebook AI Researcher, João Loula, of École Polytechnique and Facebook AI Research, and Marco Baroni of Facebook AI Research, explore recurrent neural networks’ challenges in compositional abilities and generalization.
The academic medical center of the University of Michigan is leveraging investments in artificial intelligence, machine learning and advanced analytics to unlock the value of its health data.
According to Andrew Rosenberg, MD, chief information officer for Michigan Medicine, the organization currently has 34 ongoing AI and machine leaning projects, 28 of which have principal investigators.
“There’s a lot of collaboration around these projects—as there should be for the diversity of thought and background needed to deal with complex problems—working with at least seven other U of M schools,” Rosenberg told the Machine Learning for Health Care conference on Friday in Ann Arbor, Mich. “That’s one of the powers that we enjoy.”
The University of Maryland, Baltimore (UMB) and the University of Maryland, Baltimore County (UMBC) signed an agreement Aug. 12 designed to leverage UMBC’s expertise in cybersecurity and artificial intelligence/machine learning to protect medical data and devices from cyberattacks, and also to collaborate on greater data-based medical research.
“Whereas before we would think about innovations, technological innovations, now we always think about cybersecurity as part of that, that is part of the project, with a clinical project or a research project,” said Stephen N. Davis, MBBS, FACE, MACP, UMB vice president of clinical and translational research, director of UMB’s Institute for Clinical and Translational Research (ICTR) and professor and chair, Department of Medicine at the University of Maryland School of Medicine.
An international team led by scientists at Stanford University and the Autonomous University of Barcelona finds reason to hope trees will continue to suck up carbon dioxide at generous rates through at least the end of the century. However, the study published Aug. 12 in Nature Climate Change warns that trees can only absorb a fraction of carbon dioxide in the atmosphere and their ability to do so beyond 2100 is unclear.
If the sheer amount of information that we can tap into using the internet has made the world our oyster, then the huge success of Google is a testament to how lucrative search can be in helping to light the way through that data maze.
Now, in a sign of the times, a startup called Lucidworks, which has built an AI-based engine to help individual organizations provide personalised search services for their own users, has raised $100 million in funding. Lucidworks believes its approach can produce better and more relevant results than other search services in the market, and it plans to use the funding for its next stage of growth to become, in the words of CEO Will Hayes, “the world’s next important platform.”
From Snapchat photo filters to Super Bowl crowd surveillance to identity verification in airports, facial analysis techniques have taken the industry by storm. With a multitude of applications, this branch of artificial intelligence is ever evolving – encompassing image tagging on social media, expression recognition, security, marketing, robotics and more.
Bridging industry and education, a team of researchers led by Dr. Zhangyang “Atlas” Wang, assistant professor in the Department of Computer Science and Engineering at Texas A&M University, is collaborating with MoodMe to improve the algorithms used in the company’s facial analysis and recognition programs.
The team, which includes graduate students Ziyu Jiang and Jiayi Shen, as well as undergraduates Daniel Ajisafe and Geeth Tunuguntla, will focus specifically on the techniques used to re-identify participants in video conferences to measure their emotions and attentiveness.
Trust in scientists is on the rise in the United States, according to a survey of more than 4,000 people released on 2 August.
The survey, conducted by the Pew Research Center in Washington DC, found that 86% of people in the United States have “a fair amount” to “a great deal” of confidence in scientists to act in the public interest. The level of confidence is 10% higher than levels in 2016, the first year that Pew conducted this survey (see ‘In scientists we trust’).
For respondents who have “a great deal” of confidence in scientists to act in the public interest, levels increased from 21% in 2016 to 35% in 2019. A rise in “great” confidence occurred for most of the groups included in the survey ― including the news media and elected officials ― but it was the most pronounced for scientists.
“I started the company because I had this vast swath of research,” Corso said, “and the vast majority of services that were available were focused on image-based understanding rather than video-based understanding. And in almost all instances we’ve seen, when we use a video-based model we see accuracy improvements.”
While any old off-the-shelf algorithm can recognize a car or person in an image, it takes much more savvy to make something that can, for example, identify merging behaviors at an intersection, or tell whether someone has slipped between cars to jaywalk. In each of those situations the context is important and multiple frames of video are needed to characterize the action.
“When we process data we look at the spacio-temporal volume as a whole,” said Corso. “Five, 10, 30 frames… our models figure out how far behind and forward it should look to find a robust inference.”
Conference co-chair Vipin Kumar acknowledged in an Aug. 6 interview that Anchorage is not a typical city to host discussions about high-technology innovation but also noted that the conference, known as KDD, was recently in Halifax, Nova Scotia, among meetings in Sydney, San Francisco and Beijing.
His fellow co-chair Ankur Teredesai said it was a family trip to Alaska eight years ago — the first trip with his young daughter — that largely drove him to pitch for holding KDD here. As is often the case with first-time visitors, he was taken aback by the state’s natural features.
Teredesai also noted that conference organizers wanted to move away from a solely business-driven agenda.
“It was fascinating to me to see what would happen if 1,500 to 2,000 data scientists converged on this city and shared in that spirit of the importance of the environment and climate change,” he said.
Pew Research Center; Cary Funk, Meg Hefferon, Brian Kennedy and Courtney Johnson
from
The Pew Research Center survey asked about several factors that could potentially increase – or decrease – trust in research findings and recommendations. The two steps that inspire the most confidence among members of the public are open access to data and an independent review.
Majority of Americans say they are more apt to trust research when the data is openly availableA majority of U.S. adults (57%) say they trust scientific research findings more if the researchers make their data publicly available. Another 34% say that makes no difference, and just 8% say they are less apt to trust research findings if the data is released publicly.
About half the public (52%) say they trust scientific findings more if the findings have been reviewed by an independent committee.
Menlo Park, CA September 6, starting at 8 a.m., Facebook Headquarters. “An opportunity to foster discussion and collaboration between NLP researchers in academia and industry. The event will include talks and a panel from research leaders on the latest advances in NLP technologies.” [rsvp required]
Seattle, WA August 26-30 at University of Washington. “Oceanhackweek is a 5-day learning hackathon aimed at exploring, creating and promoting effective computation and analysis workflows for large and complex oceanographic data.” [application required]
“Developers will have the opportunity to win over $60,000 in cash prizes and more. First-place winners will also receive round-trip airfare to attend the PyTorch Developer Conference, and have their projects featured in the event.” Deadline for submissions is September 16.
Error analysis — the attempt to analyze when, how, and why machine-learning models fail — is a crucial part of the development cycle: Researchers use it to suggest directions for future improvement, and practitioners make deployment decisions based on it. Since error analysis profoundly determines the direction of subsequent actions, we cannot afford it to be biased or incomplete.
IBM Research today introduced AI Explainability 360, an open source collection of state-of-the-art algorithms that use a range of techniques to explain AI model decision-making.