Data Science newsletter – November 24, 2020

Newsletter features journalism, research papers and tools/software for November 24, 2020

GROUP CURATION: N/A

 

NSF begins planning for decommissioning of Arecibo Observatory’s 305-meter telescope due to safety concerns

National Science Foundation


from

Following a review of engineering assessments that found damage to the Arecibo Observatory cannot be stabilized without risk to construction workers and staff at the facility, the U.S. National Science Foundation will begin plans to decommission the 305-meter telescope, which for 57 years has served as a world-class resource for radio astronomy, planetary, solar system and geospace research.

The decision comes after NSF evaluated multiple assessments by independent engineering companies that found the telescope structure is in danger of a catastrophic failure and its cables may no longer be capable of carrying the loads they were designed to support. Furthermore, several assessments stated that any attempts at repairs could put workers in potentially life-threatening danger. Even in the event of repairs going forward, engineers found that the structure would likely present long-term stability issues.


In Fighting The Pandemic, Policymakers Must Cite Their Sources

WNYC, On the Media


from

If the ultimate toll of the pandemic depends on trust in government, medicine, and the pharmaceutical industry, we have a long way to go. In varying degrees of reason and paranoia, there’s a lot of distrust to go around. And government itself has perhaps been the most destructive — not just by politicizing the science behind the pandemic in the White House, but by embracing public-health policies that are arbitrary, inconsistent, and largely untethered to actual science. Writing in WIRED this week, science journalist Roxanne Khamsi argued that political leaders have failed to consistently explain the science behind their policies — and that so-called “Covid theater” has only further eroded public trust. She and Bob discuss the damage that arbitrariness and a lack of transparency can do to an already-struggling effort to quash the pandemic. [audio, 13:31]


CCC Quadrennial Papers: Artificial Intelligence

Computing Community Consortium, The CCC Blog, Maddy Hunter


from

As part of the rollout of the 2020 Computing Research Associations (CRA) Quadrennial Papers, the Computing Community Consortium (CCC) is pleased to publish the final group of papers around the “Artificial Intelligence (AI)” theme, including papers on AI being deployed at the edge of the network, cooperation between AI and humans, new approaches to understanding AI’s impact on society, AI-driven simulators, and the next generation of AI. The Quadrennial Papers are intended to help inform the computing research community and those who craft science policy about opportunities in computing research to help address national priorities. This group of papers is the final installation of the CCC’s contribution, in addition to the previous themes of Broad Computer Science, Core Computer Science, and Socio-Technical Computing.


The ethical questions that haunt facial-recognition research

Nature, News Feature, Richard Van Noorden


from

Journals and researchers are under fire for controversial studies using this technology. And a Nature survey reveals that many researchers in this field think there is a problem.


Stanford researchers combine Zillow and census data to determine residential water needs

Stanford University, Stanford News


from

The gateway to more informed water use and better urban planning in your city could already be bookmarked on your computer. A new Stanford University study identifies residential water use and conservation trends by analyzing housing information available from the prominent real estate website Zillow.

The research, published Nov. 18 in Environmental Research Letters, is the first to demonstrate how new real estate data platforms can be used to provide valuable water use insights for city housing and infrastructure planning, drought management and sustainability.

“Evolving development patterns can hold the key to our success in becoming more water-wise and building long-term water security,” said study senior author Newsha Ajami.


Research Data Services in US Higher Education

Ithaka S+R, Jane Radecki and Rebecca Springer


from

As data-driven research methods proliferate and become more sophisticated across disciplines, supporting researchers who work with data is increasingly a top priority for academic institutions. However, research data services—support offerings which enable and improve data research—are currently provided in an ad hoc manner by a variety of campus units, including libraries, academic departments and institutes, labs, and IT or research computing units. And the provision of research data services varies significantly from campus to campus.

For data-driven research to thrive, stakeholders—including academic libraries, IT departments, research offices, university administrators, and academic department leadership—must have a comprehensive understanding of the current landscape of research data services. Although there is an extensive literature documenting support for research data management, sharing, and other services by academic libraries,[1] there has been much less systematic study of the research data services provided by academic institutions holistically.[2] There is an urgent need for a complete picture of the provision of research data services in US higher education.

This report presents the results of an inventory of research data services offered by US colleges and universities using a systematic web searching process. Our results represent a holistic, quantitative picture of services that support data-driven research across organizational units and institutional types.


Machine learning can fill in missing information in public health data, U of A research shows

CBC News (Canada)


from

A new machine learning tool could help fill significant gaps in Canada’s public health data, according to research released this week.

The program is a machine learning framework that finds ethnicity information and Indigenous status in Canadian health records. The program has been used to analyze the names and locations of 4.8 million people surveyed in the 1901 Canadian census using respondents’ names, spelling, phonetics and location, among other details, to predict their ethnicity.

Kai On Wong, senior data scientist at the Northern Alberta Clinical Trials and Research Centre, said he was inspired to create the program after seeing gaps in Canadian health data while working in government and for various research groups.

“If we don’t have that information, we cannot study this kind of record and we cannot tell in a consistent and timely fashion across most of these databases which ethnic groups are experiencing worse health outcomes,” said Wong, an epidemiologist at the University of Alberta.


Why finance is deploying natural language processing

MIT Sloan, Ideas Made to Matter blog, Tracy Mayor


from

Three years into his stint teaching machine learning at MIT Sloan, finance lecturer has just one complaint: It’s hard to keep up. “It’s such a fast-moving field, a lot of what’s state-of-the-art now wasn’t invented when I taught the course a year ago,” he said.

Officially titled Advanced Data Analytics and Machine Learning in Finance, the course reflects a move in finance, normally a tech-cautious industry, to embrace machine learning to help make faster, better-informed decisions.

Specifically, financial analytics firms are turning to natural language processing to parse textual data hundreds of thousands of times faster and more accurately than humans can, said Shulman, head of machine learning at Kensho. (The startup, which specializes in artificial intelligence and analytics for the finance and U.S. intelligence communities, was acquired by S&P Global in 2018.)


Chicago tech center Discovery Partners Institute gets $23.5 million

Crain's Chicago, Greg Hinz


from

After years of promises from two governors, the budding Discovery Partners Institute in the South Loop finally is getting the first installment of the state help needed for a project considered key to Chicago’s growth as a tech center.

Gov. J.B. Pritzker announced that he has released $142 million for various university research projects, including $23.5 million for design of DPI’s 500,000-square-foot headquarters in the 78 development at Roosevelt Road and the Chicago River.

Another $98 million will go the University of Illinois at Chicago for a previously announced computer design research and learning center. The remainder is for other university projects in Chicago and other parts of the state under the state’s Illinois Innovation Network.


University of Maryland, Carnegie Mellon and Facebook partner up to track coronavirus spread

Baltimore Sun, Phil Davis


from

The University of Maryland has partnered up with Carnegie Mellon University and Facebook to analyze more than 30 million surveys in an attempt to track coronavirus outbreaks.

In a news release, the University of Maryland wrote that the group issued the International COVID-19 Symptom Survey through Facebook earlier this year, “which asks individuals about their COVID-19 behaviors and resources, such as testing availability, mask usage, and symptoms.”

The university wrote that more than 30 million users from more than 200 countries have completed the survey and that global and national data visualization tools are available online for certain sets of data.


Rice University’s Secret for Containing the Coronavirus: A Student-Run Court That Prosecutes Rule Breakers

Texas Tribune, Peter Holley


from

The two undergraduates had been accused of a bold infraction of the rules: sneaking a friend from across campus into their dorm room for some late-night socializing. As recently as last spring, nobody would’ve cared. But this semester, with Rice University officials determined to prevent a coronavirus outbreak that might shut down the school, students are not only forbidden from entering dorms they don’t live in; they’re not allowed to move between floors in their own residential building. After a tipster turned them in, the two alleged offenders were forced to appear before the student-run COVID Community Court that the school launched this semester to adjudicate public health violations on campus.

Though the alleged offenders adamantly denied the accusation against them in their Zoom hearing, the evidence was “overwhelming,” according to Mel Xiao, a twenty-year-old senior from New Jersey and one of three CCC judges, selected by application, who investigated the case. “We spoke to multiple other people who lived in their [dorm], and they all confirmed it,” said Xiao, a premed student who recovered from a battle with COVID-19 in September. “That was a case where we had to assign a more serious sanction, and they also received a pretty stern warning that essentially said, ‘We know you withheld this information and weren’t transparent with us.’ ”


New Tool Helps Parents and Educators Estimate COVID-19 Infection Numbers at Their School

University of Texas at Austin, College of Natural Sciences


from

With COVID-19 cases hitting new highs across the country, a new online tool can help families and school leaders estimate how many infected people are likely to show up at a school on a given day anywhere in the United States. The free, interactive dashboard was produced by The University of Texas at Austin COVID-19 Modeling Consortium.

“It’s meant to guide families, teachers and school leadership in understanding the risks of bringing students into classrooms and onto campuses. These risks are changing through time as the prevalence of the virus rises and falls in cities around the U.S.,” said Lauren Ancel Meyers, a professor of integrative biology and director of the consortium.


New York City wants to restrict artificial intelligence in hiring

CBS News, Irina Ivanova


from

New York City is trying to rein in the use of algorithms used to screen job applicants. It’s one of the first cities in the U.S. to try to regulate what is an increasingly common — and opaque — hiring practice.

The city council is considering a bill that would require potential employers to notify job candidates about the use of these tools, referred to as “automated decision systems.” Companies would also have to complete an annual audit to make sure the technology doesn’t result in bias.

The move comes as the use of artificial intelligence in hiring skyrockets, increasingly replacing human screeners.


Time may be right for professionalizing artificial intelligence practices

ZDNet, Joe McKendrick


from

With so much riding on the performance and accuracy of artificial intelligence algorithms — from medical diagnoses to legal advice to financial planning — there have been calls for the “professionalization” of AI developers, through mechanisms such as certifications and accreditations, all the way up to government mandates. After all, it is argued, healthcare professionals, lawyers and financial advisors all require varying levels of certification, why shouldn’t the people creating the AI systems that could replace the advice of these professionals also be verified?


Census workers say pressure to end count early led to chaos

Reveal news, Byard Duncan and David Rodriguez


from

Frustrations like [Tom] Friedmann’s were common among nearly 100 census employees across the U.S. who responded to Reveal from The Center for Investigative Reporting’s ongoing callout seeking their experiences. (Reveal verified the employment of everyone quoted in this article). Facing a cascade of shifting deadlines, workers said poor communication and training, pressure to end the count early, clunky technology and haphazard management practices produced chaos on the ground as the census came to a close Oct. 15.

This year’s count was different from in years past for a number of reasons: It was the first to have been conducted with smartphone technology. It was significantly delayed by the COVID-19 pandemic. And work to follow up with households, which was supposed to take 11 weeks, ultimately was done in nine: The Trump administration abruptly decided to speed up the process in an effort to deliver population totals to the president by the bureau’s statutory deadline of Dec. 31.

In August, Commerce Secretary Wilbur Ross, who supervises the census, insisted that the truncated operation “will meet or exceed the standard for data collection set in previous decennial censuses.” But the inspector general’s office for the Department of Commerce concluded in a report released a month later that a shortened schedule “increases the risks to the accuracy of the 2020 Census.” The same report pointed out that “Senior officials at the Bureau, including the Director, did not know who ultimately made the decision to accelerate the Census schedule.”


Events



The Self-Organizing Conference on Machine Learning is returning as a 100% online event for 2020.

Twitter, Ian Goodfellow


from

Nov 30-Dec 4. It will still be small to maintain the group discussion feel. Apply at http://socml.org


Tools & Resources



How to have a difficult conversation

Psyche, Adar Cohen


from

Avoidance will only foster more conflict. Aim for a shared understanding with these techniques from an expert mediator


NEW from a big collaboration at Google: Underspecification Presents Challenges for Credibility in Modern Machine Learning

Twitter, Alexander D'Amour


from

“Explores a common failure mode when applying ML to real-world problems.”
Thread 1/14


Careers


Full-time positions outside academia

Senior Football Intelligence Manager



Manchester City FC; Manchester, England

Project Manager



Periscopic; Portland, OR

Machine Learning Engineer



Curai; Remote and Palo Alto, CA

Director of Baseball Analytics



Cincinnati Reds baseball team; Cincinnati, OH
Internships and other temporary positions

Undergraduate Research Internship – Computing



Microsoft Research; Redmond, WA; New York, NY; Cambridge, MA
Tenured and tenure track faculty positions

Tenure-track Faculty Positions



IMDEA Software Institute; Madrid, Spain

Leave a Comment

Your email address will not be published.