Data Science newsletter – April 21, 2020

Newsletter features journalism, research papers, events, tools/software, and jobs for April 21, 2020

GROUP CURATION: N/A

 
 
Data Science News



Early experience with “universal” molecular testing for Covid-19. Tested 408 people at a Boston homeless shelter

Twitter, Antonio Regalado


from

36% positive on PCR!! Most had no symptoms at time of test

so…molecular tests for all, not just thermometer readings?


Virus forced schools online, but many students didn’t follow

Associated Press, Julie Watson and Carolyn Thompson


from

During the first week that her San Diego public school was shuttered to slow the spread of the coronavirus, not one of Elise Samaniego’s students logged on to her virtual classroom.

Three weeks in, the teacher still hadn’t connected online with roughly two-thirds of the students in her third- and fourth-grade combo class at Paradise Hills Elementary. She fears the pandemic will exact a devastating toll on education in the United States, especially at low-income schools like hers.

“I do have several students below grade level, and this is just going to make it worse,” said Samaniego, who has been emailing and calling families to get her 22 students to participate.


WVU is using smart rings, apps and algorithms to identify COVID-19 infections before symptoms occur

Pittsburgh Post-Gazette, Don Templeton


from

For three years the West Virginia University Rockefeller Neuroscience Center has equipped 30,000 people with smart rings and smartphone apps to determine, before any signs of illness, whether they had influenza.

So it required only some computer algorithm adjustments to launch a pilot project in March to monitor 200 front-line healthcare professionals for COVID-19 and determine the presence of infection and potential for contagion 24 hours before flu symptoms emerge.

Within a week, 1,000 healthcare workers will be wearing the rings and using the apps with expectations that in a month participation will grow to 10,000 in West Virginia, Philadelphia, New York, Florida and elsewhere to determine how effectively the system provides early warning signs of infection.


University of Texas study says peak in COVID-19 deaths is still weeks away

KRCR, KEYE, Melanie Barden


from

The University of Texas has released a new study projecting COVID-19 deaths for all 50 states. The study shows it may be another several weeks before we see peaks in most states.

Researchers used geolocation data from cell phones to determine how social distancing will impact the mortality rate.

After simulating about 10,000 possible futures, they found an 80 percent chance the country’s peak in COVID-19 deaths will come May 7. *It’s important to note that researchers say the observed and projected numbers reflect confirmed COVID-19 deaths only.*


Machine learning could check if you’re social distancing properly at work

MIT Technology Review, Karen Hao


from

issues an alert when anyone is less than the desired distance from a colleague.

Six feet apart: On Thursday, the startup released a blog post with a new demo video showing off a new social distancing detector.


Intel Enlists Udacity to Award Degree for AI on the Edge

RTInsights, Michael Vizard


from

Today, it is critical to increase the installed base of developers that can build and deploy applications infused with AI models.

Intel and Udacity, a provider of online educational services for IT professionals, today launched a program through which they will bestow a degree to developers that learn how to employ deep learning algorithms and computer vision to create artificial intelligence (AI) applications for edge computing platforms.


What happens in September? Prepare yourselves for ‘Social Distancing U’

University Affairs magazine, Lisa Young


from

Fall 2020 is full of uncertainty. If we have managed to “flatten the curve” of COVID-19 infections, we will have gradually regained some of our freedom to move over the summer months. But, public health officials will be wary of a second wave in the fall. Although it might be better managed, and testing should be more readily available, COVID-19 will still be very much with us when the leaves start to turn colour.

To be able to bring students back, universities will have to convince their boards that the risks and costs of staying online outweigh the risks and costs of bringing the students back. They will have to convince students and parents that they have a viable plan that prioritizes health and continuity. And, most important, they will have to convince provincial health officials that they have a comprehensive plan to carry out operations while maintaining social distancing. Start planning now for Social Distancing U, opening September 2020.


The importance of being second – PLOS-wide edition

The Official PLOS Blog, Veronique Kiermer


from

Our journals which are selective for novelty — PLOS Biology, PLOS Medicine, PLOS Computational Biology, PLOS Genetics, PLOS Neglected Tropical Diseases and PLOS Pathogens — have all revised their ‘scooping’ policy to waive the novelty criteria for work submitted within six months of a similar study having been published. If your almost-completed work was ‘scooped’ by a study published today, it will still be considered novel by these journals, giving you the time to rigorously complete it and submit it within the next six months.


Amazon-owned Whole Foods is quietly tracking its employees with a heat map tool that ranks which stores are most at risk of unionizing

Business Insider, Hayley Peterson


from

Amazon-owned Whole Foods is tracking and scoring stores it deems at risk of unionizing, according to five people with knowledge of the effort and internal documents viewed by Business Insider.

The scores are based on more than two dozen metrics, including racial diversity, employee loyalty, “tipline” calls, and violations recorded by the Occupational Safety and Health Administration.


Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims

arXIv, Computer Science > Computers and Society; Miles Brundage et al.


from

With the recent wave of progress in artificial intelligence (AI) has come a growing awareness of the large-scale impacts of AI systems, and recognition that existing regulations and norms in industry and academia are insufficient to ensure responsible AI development. In order for AI developers to earn trust from system users, customers, civil society, governments, and other stakeholders that they are building AI responsibly, they will need to make verifiable claims to which they can be held accountable. Those outside of a given organization also need effective means of scrutinizing such claims. This report suggests various steps that different stakeholders can take to improve the verifiability of claims made about AI systems and their associated development processes, with a focus on providing evidence about the safety, security, fairness, and privacy protection of AI systems. We analyze ten mechanisms for this purpose–spanning institutions, software, and hardware–and make recommendations aimed at implementing, exploring, or improving those mechanisms.


Deploying more conversational chatbots

MIT News


from

The startup Posh is trying to make conversations with chatbots more natural and less maddening. It’s accomplishing this with an artificial intelligence-powered system that uses “conversational memory” to help users complete tasks.

“We noticed bots in general would take what the user said at face value, without connecting the dots of what was said before in the conversation,” says Posh co-founder and CEO Karan Kashyap ’17, SM ’17. “If you think about your conversations with humans, especially in places like banks with tellers or in customer service, what you said in the past is very important, so we focused on making bots more humanlike by giving them the ability to remember historical information in a conversation.”

Posh’s chatbots are currently used by over a dozen credit unions across voice- and text-based channels. The well-defined customer base has allowed the company to train its system on only the most relevant data, improving performance.


The Subways Seeded the Massive Coronavirus Epidemic in New York City

NBER Working Papers, Jeffrey E. Harris


from

New York City’s multitentacled subway system was a major disseminator – if not the principal transmission vehicle – of coronavirus infection during the initial takeoff of the massive epidemic that became evident throughout the city during March 2020. The near shutoff of subway ridership in Manhattan – down by over 90 percent at the end of March – correlates strongly with the substantial increase in the doubling time of new cases in this borough. Maps of subway station turnstile entries, superimposed upon zip code-level maps of reported coronavirus incidence, are strongly consistent with subway-facilitated disease propagation. Local train lines appear to have a higher propensity to transmit infection than express lines. Reciprocal seeding of infection appears to be the best explanation for the emergence of a single hotspot in Midtown West in Manhattan. Bus hubs may have served as secondary transmission routes out to the periphery of the city.


Vaccines are ‘long shots’, coronavirus will be no different

BBC Science Focus Magazine, Alexander McNamara


from

Vaccines are “long shots” and people should not rely on the swift development of one for COVID-19, the Government’s chief scientific adviser has warned.

A group of Oxford University researchers will begin clinical trials for a coronavirus vaccine next week but Sir Patrick Vallance says expectations need to be tempered.

Writing in The Guardian, Sir Patrick wrote: “All new vaccines that come into development are long shots; only some end up being successful, and the whole process requires experimentation.”


More than $14 million in research grants awarded for health technology solutions focused on heart and brain health, including special projects related to COVID-19 and CVD

American Heart Association


from

Research teams at Cincinnati Children’s Hospital, The Johns Hopkins University, Stanford University School of Medicine and the University of Michigan will receive $2.5 million each for their individual projects aimed at reducing health care disparities, empowering people to better manage their health and wellness, and enhancing patient/provider connectivity. Together, they’ll also receive $4 million to work collectively on at least one highly impactful project and form a national Health Technology Research Collaborative. The Collaborative may ultimately serve as an American Heart Association research ‘think tank’ to assist with identifying, creating, testing and bringing to scale future innovative health technologies.

In addition to the original awards, each team can also apply for supplemental research grants of up to $200,000 for rapid action projects to develop technology solutions to address the COVID-19 pandemic. The projects could possibly provide aid for health care systems, doctors or care providers, first responders, patients or consumers. The supplemental grants are part of the American Heart Association’s $2.5 million commitment to research efforts to better understand this unique coronavirus and its interaction with the body’s cardiovascular and cerebrovascular systems.

“The peer review committee has assembled an exceptional network to move this work forward and I want to recognize the dedication and commitment of that panel of many renowned experts,” Harrington said. “The Association uses an intense, multi-stage review process in selecting the centers for our focused research networks and we’re very appreciative of the committee members who lend their time and expertise to this critical process.”


Introducing Our New Executive Director

The Carpentries


from

We are thrilled to announce that Dr. Kari L. Jordan is the new Executive Director of The Carpentries, effective 16 April, 2020. This announcement comes after a rigorous international search conducted by Marcum LLP and members of the search committee (Karen Cranston, Amy Hodge and Elizabeth Wickes) appointed by the Executive Council of The Carpentries.

 
Deadlines



neuromatch 2.0

Online May 25-27. “An unconference in Computational Neuroscience” Deadline for submissions is May 20.

Paul Evan Peters Fellowship

“The Paul Evan Peters Fellowship was established to honor and perpetuate the memory of Paul Evan Peters (1947-1996), founding executive director of the Coalition for Networked Information (CNI). The fellowship will assist students pursuing graduate studies in the information sciences, librarianship, or closely related field, that advance the frontiers of digital information and technology.” Deadline for applications is May 24.
 
Tools & Resources



Open Data on COVID-19 Now Available in Data Observatory

CARTO Blog; Javier de la Torre, Javier Pérez Trufero, and Alejandro Polvillo


from

In order to empower these professionals with more tools to analyze and provide a better response to the consequences of this pandemic, CARTO has been making its platform available to public and private sector organizations doing research on the coronavirus outbreak via our grants program. To date hundreds of grants have been approved in 35 countries. In addition, our team of data scientists and engineers have been actively working to build tools and providing analysis to several administrations worldwide.

Amongst these efforts, we have experienced first hand how important it is to have access to ready-to-use and up-to-date datasets, so data scientists can spend the majority of their time building models and interpreting the results before they become obsolete in this rapidly changing environment.


COVID-19 Collaboration Platform (CovidCP) is a platform for collaborating on RCTs to find treatments more efficiently and reliably.

Twitter, Betsy Ogburn


from

Clinical researchers, please submit your draft or finalized RCT protocols at https://covidcp.org so that other researchers can collaborate!


[keras.io] transfer_learning.ipynb

Francois Chollet


from

“Transfer learning consists of taking features learned on one problem, and leveraging them on a new, similar problem. For instance, features from a model that has learned to identify racoons may be useful to kick-start a model meant to identify tanukis.”


Pandas 100 tricks

Kaggle, Kevin Markham


from

“This kernel is a compilation of tricks of pandas published by Kevin Markham weekly.”

 
Careers


Full-time positions outside academia

Senior Epidemiologist/Study Director



Westat; Atlanta, GA
Internships and other temporary positions

Text Mining Student Specialist



University of Washington, eScience Institute; Seattle, WA

Leave a Comment

Your email address will not be published.