Data Science newsletter – March 12, 2021

The Pittsburgh Penguins have named Katerina Wu as data scientist in the team’s hockey operations department, it was announced today by general manager Ron Hextall.

Wu, 22, will design and implement new statistics to evaluate player and team performance. She will report to Sam Ventura, the Penguins’ director of hockey operations and hockey research, while also working closely with Nick Citrone, who has been promoted to senior data scientist of hockey and business operations.

Scientific papers are the backbone of a research community and the citation of those papers sparks conversation in a given field. This cycle of publication and citation leads to new knowledge, but what happens when implicit discrimination in a field leads to papers by minority scholars being cited less often than their counterparts? A new team of researchers has come together to ask this question and dig into the numbers of gender and racial bias in neuroscience.

The team members include physicist and neuroscientist Danielle Bassett, J. Peter Skirkanich Professor of Bioengineering at the University of Pennsylvania, with secondary appointments in the Departments of Neurology and Psychiatry in Penn’s Perelman School of Medicine, statistician Jordan Dworkin, then a graduate student in Penn Medicine’s Department of Biostatistics, Epidemiology and Bioinformatics, and ethicist Perry Zurn, an Assistant Professor of Philosophy at American University.

Their study on gender bias, which recently appeared in Nature Neuroscience, reports on the extent and drivers of gender imbalance in neuroscience reference lists. The team has also published a perspective paper in Neuron that makes practical recommendations for improving awareness of this issue and correcting for biases.

New Year’s Eve 2019: Ian Lipkin, a famed Columbia University epidemiologist, is having dinner with his wife and a fellow scientist. He gets a confidential phone call from a highly placed source in China: There’s a cluster of pneumonia-like illnesses in the city of Wuhan caused by a novel coronavirus. The source says it’s not that big a deal: It doesn’t look very transmissible.

“I was told not to worry about it,” Lipkin recalls.

It was something to worry about.

That virus, later named SARS-CoV-2, would slowly reveal its secrets — and proceed to shut down much of the planet, killing more than 2.6 million people in the most disruptive global health disaster since the influenza pandemic of 1918.

The world is now confronting a growing number of coronavirus variants that threaten to slow or undo our vaccine progress. In recent months, it’s become clear that the virus is mutating in ways that make it more transmissible and resistant to vaccines, and possibly more deadly. It’s also clear that, at least in the United States, there is no organized system for tracking the spread or emergence of variants. As [Adriana] Heguy sees it, the U.S. has more than enough genome-sequencing expertise and capacity; the problem is focus. “Efforts in the U.S. have been totally scattered,” she said. “There’s no mandate to do it in a timely fashion. The government is kind of like, Let us know if you find something.” Funding has also been a major constraint. “It boils down to money,” Heguy said. “With money, I could hire a technician, another scientist, get the reagents and supplies I need.” Because of their better-organized efforts, other countries have been more successful in identifying new versions of the virus: “The reason the U.K. variant was identified in the U.K. is that the U.K. has a good system for identifying variants.” The U.K. has, for months, sequenced at least ten per cent of its positive tests. “If you’re doing ten per cent, you’re not going to miss things that matter,” Heguy said. “If a variant becomes prevalent, you’ll catch it.”

The insects seemed to be disappearing in areas where fall temperatures had risen significantly more than summer temperatures over the past several decades, as in the U.S. Southwest. The warmer weather may disrupt the butterfly’s breeding cycle or negatively affect the plants they depend on, Forister speculates.

“[The new study] did a good job of working with what is inherently messy data,” says Scott Hoffman Black, an ecologist and executive director of the Xerces Society for Invertebrate Conservation. Given that butterflies are key pollinators, such declines portend bigger problems for plants and even whole ecosystems, Forister adds. “The climate effects will almost certainly affect many other insects, including bees.”

Another smart city project tied to Sidewalk Labs has come to an early close.

RedTail Media reporter Kate Kaye first reported in late February that Portland Metro had severed ties from its smart city work with Sidewalk Labs spin-off company Replica.

Portland Metro began working with Replica, which separated from Sidewalk Labs in September 2019, to provide insight into how local residents moved throughout the Portland, OR region, kicking off the project in April 2019.

But the partnership ended roughly two years later due to disagreements over data privacy, a sticking point that has become increasingly common in smart city projects, and an issue that experts say local leaders and their private partners can take note of to learn from and improve future collaborations.

Change Healthcare’s dataset – diagnoses, prescriptions, and SDOH information – is drawn from across the U.S. healthcare system. Each DSaaS instance is dedicated to the specific client, the company says, with the opportunity to add other datasets, analytic tools or methods as the project or use case requires.

THE LARGER TREND

Several high-profile customers are already making use of the DSaaS offering. Duke University School of Medicine has leveraged the cloud service to compare differences in COVID-19 disease progression depending on pre-existing conditions and various interventions for different ethnic and socio-economic subgroups.

I have seen so many posts on social media about how great pytorch is and, in one latest tweet, ‘boomers’ use tensorflow … It doesn’t make sense to me and I see it as being incredibly powerful and widely used in research and industry. Should I be jumping ship? What is the actual difference and why is one favoured over the other? I have only used tensorflow and although I have been using it for a number of years now, still am learning. Should I be switching? Learning both? I’m not sure this post will answer my question but I would like to hear your honest opinion why you use one over the other or when you choose to use one instead of the other.

EDIT: thank you all for your responses. I honestly did not expect to get this much information and I will definitely be taking a harder look at Pytorch and maybe trying it in my next project. For those of you in industry, do you see tensorflow used more or Pytorch in a production type implementation? My work uses tensorflow and I have heard it is used more outside of academia – mixed maybe at this point?

Robust information systems are not free. They require time, attention and money — but they can pay for themselves over and over again in better decisions taken, and better democratic accountability after the fact. … It isn’t cheap to build the systems that show you what’s coming at you. But failing to build them? That’s far more expensive.

Elite schools breed entitlement, entrench inequality—and then pretend to be engines of social change.

JMU biology professors Bisi Velayudhan and Corey Cleland received $359,768 from the National Science Foundation to run a 10-week research program at JMU for the next three summers.

As a part of the NSF’s Research Experiences for Undergraduates Program, each cycle will accept up to 10 community college students who’ll live on campus and conduct research with faculty mentors. The deadline to apply for this summer’s program is March 15, and the session will be held May 23 through July 30.

Research from the College Board found that 51% of all undergraduates enrolled at public two-year colleges identify as a race or ethnicity other than white. According to a survey by USA Today, 70% of the STEM field is white.

In a new study, the team developed and applied artificial intelligence (AI), or machine-learning, algorithms to physiological data — including respiratory rate, blood pressure, heart rate, body temperature and oxygen levels — from patients with chronic pain from sickle cell disease. Not only did the researchers’ approach outperform baseline models to estimate subjective pain levels, it also detected changes in pain and atypical pain fluctuations.

Using artificial intelligence, scientists can now rapidly generate photorealistic color 3D holograms even on a smartphone. And according to a new study, this new technology could find use in virtual reality (VR) and augmented reality (AR) headsets and other applications.

Looking for a research paper but can’t find a copy in your library’s catalog or popular search engines? Give Internet Archive Scholar a try! We might have a PDF from a “vanished” Open Access publisher in our web archive, an author’s pre-publication manuscript from their archived faculty webpage, or a digitized microfilm version of an older publication.

We hope Internet Archive Scholar will aid researchers and librarians looking for specific open access papers that may not be otherwise available to them.

Artificial intelligence technologies are quickly evolving and changing every aspect of industry in the United States and globally. Artificial intelligence enables autonomy by robotic mobility and control learned through examples and computational decision-making and estimation from data using past training data experience. It has the ability to process large amounts of data much faster and make predictions more accurately than humanly possible.

To rapidly advance the field of artificial intelligence and autonomy, Florida Atlantic University’s College of Engineering and Computer Science recently unveiled its “Center for Connected Autonomy and Artificial Intelligence” (CCA-AI), a cutting-edge center designed to accelerate the development of innovative artificial intelligence and autonomy solutions.

Hamilton, Bermuda October 18-20. “Conceived by Zak Kohane, chair of the department of biomedical informatics in the Blavatnik Institute at Harvard Medical School, the symposium will bring together the brightest minds from academia and industry in the fields of computer science, AI, clinical medicine, and healthcare.” [save the date]

Online April 12-16. “Eleven years after Geoffrey Hinton couldn’t get a free sample from Nvidia, the Touring Award winner will join his comrades Yoshua Bengio and Yann LeCun at the 2021 GTC conference hosted by Nvidia as a headline speaker.” [registration required]