Data Science newsletter – October 30, 2020

GROUP CURATION: N/A

Health officials in Michigan issued a stay-in-place order for undergraduate students at the University of Michigan, effective immediately, as new confirmed coronavirus cases spike. One big exception: Football.

The public health directive does not apply to varsity student athletes, who are exempt and allowed to take part in practices and competitions.

Washtenaw County, which includes University of Michigan’s campus in Ann Arbor, announced the order Tuesday. It is effective through 7 a.m. on Nov. 3.

#DataScienceEducation is fast approaching public value failure status. We need to “[create]a taxonomy capable of deploying data science that reflects the values of the communities they aim to serve.” Starting w/ PV4 + PV5

Amazon’s Alexa is getting better at recognizing who’s speaking and what they’re speaking about, understanding words through on-device techniques, and leveraging models trained without needing human review. That’s according to automatic speech recognition head Shehzad Mevawalla, who spoke with VentureBeat ahead of a keynote address at this year’s Interspeech conference.

Alexa is now running “full-capability” speech recognition on-device, after previously relying on models many gigabytes in size that required huge amounts of memory and ran on servers in the cloud. That change is because of a move to end-to-end models, Mevawalla said, or AI models that take acoustic speech signals as input and directly output transcribed speech. Alexa’s previous speech recognizers had specialized components that processed inputs in sequence, such as an acoustic model and a language model.

With a background in all facets of digital and social media marketing and a proven track record in achieving revenue and traffic gains, Eric Ames is leading way for new publisher Lines.com to capitalize on the sports betting media revolution and impact gambling and fantasy sports through up-to-the-minute insights, information and content.

In a chat with Hashtag Sports, Eric shares how his small team is leveraging content automation to create up to 30 new pieces of content per day, the importance of longtail content in the publisher’s SEO success, and why the pandemic hasn’t changed the site’s launch strategy.

Modern health care has been reinvigorated by the widespread adoption of artificial intelligence. From speeding image analysis for radiology to advancing precision medicine for personalized care, AI has countless applications, but can it rise to the challenge in the fight against Covid-19?

Researchers from the Abdul Latif Jameel Clinic for Machine Learning in Health (Jameel Clinic), now housed within the MIT Stephen A. Schwarzman College of Computing, say the ongoing public health crisis provides ample opportunities for leveraging AI technologies, such as accelerating the search for effective therapeutics and drugs that can treat the disease, and are actively working to translate this potential to success.

COVID-19 testing is now a part of daily life for on-campus students, faculty and staff at the College. Moving through the Greylock parking garage and now the Towne Field House, students have become accustomed to the highly organized process. They insert a swab into a test tube, seal the lid and move on with their day. Within 24 hours, their results arrive.

Yet there’s so much more operating beneath the surface, from staff coordination to transportation to the testing process itself. The process that takes students under five minutes to complete has been in development since July.

Rita Coppolla-Wallace, normally the College’s executive director of design and construction, was charged with the task of designing a comprehensive testing site in July of this year. The first tests were planned for Aug. 17, leaving Coppola-Wallace less than a month to organize a testing center. Ultimately, she chose Greylock parking garage for its location and drive-through accessibility, taking into consideration what would make people feel most safe.

The Defense Department’s Joint Artificial Intelligence Center (JAIC) plans to expedite funding to adapt commercially-available artificial intelligence technologies for government use.

Officials at the JAIC issued what’s known as a commercial solutions opening (CSO) on Oct. 27 seeking information on commercial A.I. prototypes in more than a half-dozen key fields. CSOs are an acquisition method similar to other transaction agreements (OTAs): they enable a handful of federal agencies to negotiate binding contracts for access to critical technologies quickly without having to go through the traditional procurement process.

Until this year, Akiko Iwasaki had never had tubes of human blood delivered to her lab. “We were mostly working with mouse models,” says the Yale University immunologist, who speaks precisely and thoughtfully. “We used to look at the data and contemplate it.” Then COVID-19 struck, and such unhurried musings flew out the window. In a matter of weeks, Iwasaki overhauled her research to launch a slew of studies on how the new virus, SARS-CoV-2, takes its toll on patients. She and her nearly two dozen lab members know their discoveries could impact people falling sick right now. “Every minute counts.”

In the months since, she has produced a string of high-profile papers in which she has redirected her expertise in the immune system, honed in mice, to questions such as why men are more likely than women to fare poorly if infected and how immune responses in hospitalized patients can help predict their prognosis. Now, she is turning her attention to long-haulers, people who suffer a bout with the virus and don’t fully recover.

Iwasaki has had decades of practice adapting to new circumstances.

As AI plays a greater role in our lives both at work and at home, questions arise. How willing are we to trust AI systems? And what are our expectations for how AI should be deployed and managed?

To find out, we surveyed a nationally representative sample of more than 2,500 Australians in June and July 2020. Our report, produced with KPMG and led by Nicole Gillespie, shows Australians on the whole don’t know a lot about how AI is used, have little trust in AI systems, and believe it should be carefully regulated.

We are excited to announce that we are collaborating with U.S. universities that serve significant populations of Black and Latino students so that we can co-teach and fund graduate-level online deep learning courses. We developed and co-taught the first course last spring alongside our longtime partner Georgia Tech, and will expand this program in 2021 to additional institutions and universities with highly diverse populations, including minority-serving institutions.

“Facebook’s involvement and interest in increasing underrepresented communities in the artificial intelligence field and investment of resources convinced me to commit to teaching this course. Facebook’s resources have significantly reduced the burden of creating and teaching a high-quality course,” said Zsolt Kira, associate director of the Machine Learning Center at Georgia Tech and an assistant professor in the School of Interactive Computing.

In a new international cross disciplinary study, researchers have used artificial intelligence to analyse large amounts of historical photos from WW2. Among other things, the study shows that artificial intelligence can recognise the identity of photographers based on the content of photos taken by them.

Researchers from the Pritzker School of Molecular Engineering at the University of Chicago, however, announced they have finally found a way to design polymers by combining modeling and machine learning.

By computationally constructing nearly 2,000 hypothetical polymers, they were able to create a large enough database to train a neural network to understand which polymer properties arise from different molecular sequences.

“We show that the problem is tractable,” said Juan de Pablo, the Liew Family Professor of Molecular Engineering, who led the research. “Now that we have established this foundation and have shown that it can be done, we can really move forward in using this framework to design polymers with specific properties.”

Due to COVID-19, a majority of classes have moved to online operations with professors choosing to upload lecture videos instead of hosting in-person classes. In Computer Science and Engineering, instructors are using a software called ClassTranscribe developed by University professors to advance the online learning experience.

“My project allows students to easily go from one video to the next,” Lawrence Angrave, teaching professor and co-developer of ClassTranscribe, said. “It also allows students to search across the whole course.”

This semester due to COVID-19, Professor Angrave encountered IT challenges when ClassTranscribe had to expand suddenly to support thousands of students.

The challenge of disentangling political communication processes and their effects has grown with the complexity of the new political information environment. But so have scientists’ toolsets and capacities to better study and understand them. We map the challenges and opportunities of developing, synthesizing, and applying data collection and analysis techniques relying primarily on computational methods and tools to answer substantive theory-driven questions in the field of political communication. We foreground the theoretical, empirical, and institutional opportunities and challenges of Computational Communication Science (CCS) that are relevant to the political communication community. We also assess understandings of CCS and highlight challenges associated with data and resource requirements, as well as those connected with the theory and semantics of digital signals. With an eye to existing practices, we elaborate on the key role of infrastructures, academic institutions, ethics, and training in computational methods. Finally, we present the six full articles and two forum contributions of this special issue illustrating methodological innovation, as well as the theoretical, practical, and institutional relevance and challenges for realizing the potential of computational methods in political communication..

Of the 45 positive COVID-19 cases recorded among undergraduate and graduate students on Stanford’s campus, 29 are student-athletes. Thirteen of those are football players, according to Stanford Athletics Communications Director Scott Swegan. Twelve students are currently in isolation on campus.

Swegan told The Daily that student-athletes participating in contact practices receive “polymerase chain reaction (PCR) testing three times per week,” in addition to the football program’s daily antigen testing. This combined testing strategy follows “the County of Santa Clara’s most recent guidance for collegiate athletics.”

Online November 17, starting at 7 p.m. Eastern time. “During this dynamic event, we’ll announce the 2020 #CSforALL Commitments by our member community and highlight promising new work in response to the impacts of COVID-19 and racial injustice on the computer science education movement. This event will feature celebrity presenters, CS education influencers, a high impact keynote, and more.” [free, registration required]

“Computer scientists need to do more to understand and account for the underlying societal contexts in which these technologies are developed and deployed.”

“Here at Google, we started to lay the foundations for what this approach might look like. In a recent paper co-written by DeepMind, Google AI, and our Trust & Safety team, we argue that considering these societal contexts requires embracing the fact that they are dynamic, complex, non-linear, adaptive systems governed by hard-to-see feedback mechanisms. We all participate in these systems, but no individual person or algorithm can see them in their entirety.”