Borealis AI, an RBC Institute for Research, has expanded its artificial intelligence (AI) research team in Canada by adding Professor Pascal Poupart of the University of Waterloo and Professor Marcus Brubaker of York University as senior researchers.
Poupart is well-known for an extensive background in the development of algorithms related to machine learning, natural language processing and telecommunication networks. At the University of Waterloo, Poupart has worked with companies like Google, Intel, Huawei and Ford.
The Center for Data Science at the University of Massachusetts Amherst has been selected by IBM to join the company’s AI Horizons Network, an international consortium of leading universities working with IBM to develop technologies needed to help fulfill the promise of artificial intelligence (AI). The collaboration builds on the international reputation of the University’s College of Information and Computer Sciences as a research leader in artificial intelligence and information retrieval.
IBM will work side by side with UMass Amherst, where researchers from both entities will “develop new methods to extract semantic meaning from text, and discover broadly useful logical implications using deep neural networks,” says UMass Amherst Center for Data Science Director Andrew McCallum. McCallum is a leading expert in statistical machine learning applied to text, including information extraction, graphical models, and deep learning.
Since Zhengzhou railway police started using the eyewear earlier this year, they have identified seven people suspected of crimes ranging from human trafficking to hit-and-run accidents, according to the report.
In a similar move, train stations in major Chinese cities including Zhengzhou introduced a “face-swiping” check-in service during the lunar new year holiday in 2017. Similar to e-passport services at airports, small kiosks at boarding areas use facial recognition technology to scan passengers and their travel documents in just a few seconds.
China has pursued an ambitious plan to develop its AI sector in recent years, with police departments across the country implementing facial recognition technology. Shanghai has used it to identify and fine traffic violators, while in coastal Qingdao, facial recognition helped police arrest dozens of suspected criminals at the city’s famous beer festival.
Every NHS trust assessed for cyber security vulnerabilities has failed to meet the standard required, civil servants have said for the first time.
In a parliamentary hearing on the WannaCry attack which disrupted parts of the NHS last year, Department of Health (DoH) officials said all 200 trusts had failed, despite increases in security provision.
The WannaCry attack that began on 12 May is believed to have infected machines at 81 health trusts – nearly a third of the 236 NHS trusts in England – plus computers at almost 600 GP surgeries, according to a National Audit Office (NAO) report released in October.
University Data Science News
John Hennessy, former President of Stanford University, is replacing Eric Schmidt as board chairman at Alphabet. Hennessy is also a director at the Gordon and Betty Moore Foundation that provides funding for this newsletter.
Eric Schmidt will joinMIT as a visiting innovation fellow.
University of Michigan now has a masters in data science. It draws together four departments – biostatistics, computer science, statistics, and the information school. They are now accepting applications for the inaugural class starting Fall 2018.
UC-Berkeley‘s Division of Data Science has partnered with EdX to make Foundations of Data Science available online as a certificate program. Configured as a series of 3 five-week classes, no prior experience with statistics or programming is required.
The University of Virginia has a new masters in business analytics. They already have a masters in data science, but this new program will expect less coding and modeling, instead emphasizing “proficiency in business challenges facing organizations.”
The University of Southampton has opened a Centre for Machine Intelligence within their Department of Electronics and Computer Science. This is a less interdisciplinary approach than similar data science centers at other universities.
University of Alberta in Edmonton is one of the Canadian universities getting heavily involved in machine learning, receiving $125m from the Canadian federal government (along with two other schools) and being the lucky neighbor of DeepMind’s latest new office.
Also in Canadian AI news, there’s a gushy love piece on Geoff Hinton and all the students he has set forth into the wider world of neural nets.
University of Washington oceanographers are using SeaFlow, a new data gathering instrument that continuously profiles microbial populations, in sea water during weeks-long research cruises. The idea of a research cruise sounds great except the part where colleagues have to sleep on a small boat. For days.
PLOS Biology announced it will publish what it’s calling “complementary papers” that bolster recently published studies with similar methods and findings. This is a good way to combat the replication crisis and encourage scientists to conduct longer-term projects without a fear of getting scooped.
UMass-Amherst’s Center for Data Science is partnering with IBM through the IBM AI Horizons Network. Their collaboration will revolve around extracting semantic meaning from text. One of the reasons UMass-Amherst was selected is the size of their compute cluster: 800 GPUs thanks to a $5m grant from the Massachusetts Tech Collaborative.
New England colleges are anticipating have to scramble to fill their seats in 2025 due to a decrease in the number of children born following the 2008 recession. To compete for students, some will offer more financial aid, though that is unlikely to be an effective strategy in aggregate.
The Koch Foundation is pouring $77m annually into 300 US universities, especially those that espouse the conservative economic policies they favor. Jane Mayer wrote an entire, enraging book Dark Money about the use of universities as a trusted mouthpiece from which to propagate libertarian fiscal policy.
Virginia Tech professor Randall Murch is working with the Department of Defenseon cyberbiosecurity threats. Threats arise from “the biotechnology industry’s and other life science and medical fields’ increased reliance on computer-controlled instruments and networks” … “leaving biological data, critical instrumentation and facility operations vulnerable to cyber-based attacks.”
Kopernio is a free web-based software that enables one-click access to academic journal articles for more than 9m researchers and scientists worldwide and across a wide range of academic and researcher workflows. It intelligently activates on over 20,000 journal websites, databases, and search platforms such as PubMed and Google Scholar. Specifically, Kopernio automatically builds “smart links”, which are aware of the researcher’s context, straight to the journal article PDF, delivering content via the users institutional library subscriptions, or open sources such as pre-print servers and institutional repositories. These smart links are then placed behind a single button, so for academics and researchers who use Kopernio it’s just one click to the PDF, saving them significant amounts of time and improving their efficiency.
“I can buy my music or groceries on the internet with just one click. However, researchers who need journal papers are forced through a series of archaic login forms, redirects, and clicks, and still only get the PDF if they are lucky. Kopernio is wrapping this experience into one click, improving the efficiency of researchers around the world by making millions of research papers more accessible. We are very grateful for the support, and are proud to be able to partner with Innovate UK on this journey”, says Jan Reichelt, Kopernio CEO.
Internet radio company Pandora said Wednesday that it will implement a broad restructuring that includes cutting jobs and that it plans to expand its operations in Atlanta and away from its Oakland headquarters.
In a filing with the Securities and Exchange Commission, Pandora said it would cut 5 percent of its workforce, which would amount to 125 jobs from its current base of about 2,500 employees.
Pandora said in a statement that the company will put more of its focus on ad-tech and audience development efforts, and it expects the moves to save $45 million annually.
Applying social science approaches to conservation research is growing in popularity, but as wildlife biologists step outside their quantitative world, they can find themselves wandering unfamiliar territory.
Nibedita Mukherjee, a postdoctoral researcher at the University of Exeter and visiting researcher at the University of Cambridge, found that out herself. For her doctoral thesis at the University of Exeter, she applied a social science method called the Delphi technique — an expert consultation method that originated from Cold War bombing calculations — to wildlife biology. But she discovered she had few guidelines to follow.
“We are taught to count birds and trees in the field, so inadvertently we may be trying to do a natural science version of social science, and that leads to the improper application of techniques or bad reporting,” Mukherjee said.
That’s why she and an international group of peers started compiling papers that could help make social science application easier and more effective for conservation scientists. From her own research, Mukherjee said, she saw the role that qualitative social science methods could play in ecology and conservation, but wildlife biologists could use some guidance in putting them to work.
More than 60 researchers and technologists are running for federal office in 2018 as part of a historic wave of candidates with science backgrounds launching campaigns.
At least 200 candidates with previous careers in science, technology, engineering and math announced bids for some of the nation’s roughly 7,000 state legislature seats as of Jan. 31, according to data that 314 Action, a political action committee, shared exclusively with HuffPost.
The group, which launched in 2014 to help scientists run for office, said it is talking with 500 more people and is pressing about half of them to run. An additional 200 such candidates are running for school boards.
Although colleges for years now have made at least some effort to diversify their campuses, the country’s changing demographics will soon give them no choice.
The nation’s high school population is becoming increasingly diverse and increasingly unable to afford high tuition prices. Additionally, experts predict a major drop in the number of high school graduates overall after the year 2025 — especially in New England — because people have had fewer babies since the 2008 economic recession. As a result, local colleges will have to work harder to bring students to campus and offer them significantly more financial assistance. And some of them, experts predict, will find this a daunting new calculus, leading to more college mergers and even closures.
“Institutions in places like Massachusetts and New York and Illinois are going to be really challenged to maintain enrollments,” said Joseph Garcia, president of the Western Interstate Commission for Higher Education, whose research on this topic is the industry gold standard. “There are just not going to be enough wealthy, full-paying students to go around.”
So how can we encourage people to accept the use of algorithms when they will provide a superior or safer outcome? Once self-driving cars become safer than human drivers (if they’re not already), the unwillingness to use them will lead to more dangerous roads and deaths. Similarly, the failure to use superior decision-making tools in hospitals, schools, and other high-stakes environments is already costly.
One interesting possibility comes from an experiment by Berkeley Dietvorst, Joseph Simmons, and Cade Massey. The experimental subjects were given an algorithm to assist in predicting the percentile ranking of a student in their class. One set of experimental subjects were given the option of using only an algorithm or only their own judgement. A second group were given the choice of being able to use an algorithm for which they could adjust the result or their own judgement. The adjustment mechanism allowed the subjects to shift the algorithm’s prediction up or down by 10, 5, or 2 percentage points.
Giving subjects a role in the decision through the adjustment mechanism made them more likely to use the algorithm than those who simply had to accept its recommendation
Massachusetts General Hospital, Proto magazine, Marcia Lerner and Charles Slack
Thanks in no small part to investments in researching cancer cells, gene sequencing, immunology and dozens of other fields, an American child born today can expect to live 30 years longer than one born in 1900. “Now we can take advantage of diagnostics, devices, drugs and behavioral interventions that were unimaginable when I was in training 30 years ago,” says Christopher Austin, a neurologist and the director of NIH’s National Center for Advancing Translational Sciences (NCATS).
Early in the new century, however, the broad consensus that medical research was worth every penny began to unravel. A combination of budget cuts and modest but steady inflation led to an almost 25% reduction in the amount of research that was funded by NIH grants between 2003 and 2015. At the same time, competition for academic jobs like the one Hotez found at the outset of his research career became increasingly fierce. For the more than 40,000 new Ph.D.s in science and engineering earned each year in the United States, there are just 3,000 full-time jobs available at U.S. universities. According to a 2015 National Science Foundation survey, six in 10 newly minted Ph.D.s in the life sciences had yet to receive commitments for postdoctoral positions or other employment in their fields.
The Koch Foundation’s giving is notable in part for its vast, and almost unrivaled, scope: it now gives money to more than 300 colleges, from elite universities to rural community colleges and state flagships. In 2016, tax filings show, the foundation increased its giving by a sharp 75%, to a total of $77 million. In 2017, a spokesperson told BuzzFeed News, it gave more than $90 million, with some $60 million going to higher education — a number it expects to rise again substantially in 2018.
“We don’t have a number or a cap, at least as far as I know,” said John Hardin, who is in charge of university giving at the foundation. “You hear the stories from students and scholars that wouldn’t be able to do their work without funding, and you think — yeah, bring it on. More and more.”
Much of the foundation’s money has gone toward conservative-leaning institutions and economics-related causes: George Mason University’s Mercatus Center, a deeply influential think tank on the free market; an economic analysis institute at the University of Utah; the business school at Utah State University.
Overall, we think this move is great news. It signals that Apple, the most valuable company in the world, is making investments into consumer-driven healthcare. As long-time and sincere advocates for people having access to their complete medical histories, we see Apple’s move into healthcare as a great signal to everyone (health systems, patients, and providers) that they’re investing time and resources into building infrastructure for patients to get transparency into their healthcare information. Though this isn’t a direct move to solve interoperability, it’s a stepping stone toward making patient-owned data a reality.
While this development is encouraging, there have been many attempts and failed experiments with patient portals. The underlying issue is that patients typically don’t care about their healthcare information outside an episode of treatment or an acute setting.
A computational pathology startup has licensed technology from Memorial Sloan Kettering Cancer Center to apply machine learning algorithms to a library of pathology slides to advance oncology clinical decision support. In a phone interview, Paige.AI Cofounder and Chief Science Officer Thomas Fuchs said the funds raised from a $25 million Series A round will be used towards expanding the Memorial Sloan Kettering spinoff’s staff from five to 35 this year.
Empatica, a Massachusetts Institute of Technology spinoff, received FDA clearance for a wristworn device that uses machine learning to alert people with epilepsy and their caregivers of a convulsive seizure and track their duration and frequency.
Epilepsy affects a least 2.2 million people, according to data from the Epilepsy Foundation.
Empatica’s Emrace device assesses multiple indicators of a seizure, including electrodermal activity, a signal associated with fight or flight response that’s used by stress researchers to quantify physiological changes related to sympathetic nervous system activity, the company statement noted.
Scores of companies are speculating on a world of electric cars and self-driving vehicles, prophesying the end of car ownership, prototyping autonomous transit pods and even drawing up new plans for gas station real estate. There’s good reason for it: Ridesharing startups like Uber and Lyft have shaken up urban transportation in the last few years, at the same time as residents and city leaders are confronted with the decay of public infrastructure, including the ailing New York City subway system.
That big data will occupy the defining role in the future of mobility is already evident at the startup Teralytics. Currently operating on three continents, the company harnesses vast amounts of telecommunications data for its projects for clients in the public and private sectors. The idea is to use these detailed insights on how people move to inform everything from multimillion-dollar public transit upgrades to late-night ridesharing prices.
“In 2008 Alex White, Samir Rayani, and I set out to answer a question: how does a band become famous? We spent the next decade building the music analytics company Next Big Sound in pursuit of the answer. It was an insane, exciting, wouldn’t-trade-it-for-the-world experience where we tracked data on hundreds of thousands of artists, established a track record of predicting break-out stars, published industry-redefining research, and supplied analytics and insights to the vast majority of the music industry, from individual artists to the major labels. We learned a ton and had so much fun.