Data Science newsletter – February 1, 2022

Newsletter features journalism, research papers and tools/software for February 1, 2022

 

MIT Energy Initiative launches the Future Energy Systems Center

MIT News, MIT Energy Initiative


from

The MIT Energy Initiative (MITEI) has launched a new research consortium — the Future Energy Systems Center — to address the climate crisis and the role energy systems can play in solving it. This integrated effort engages researchers from across all of MIT to help the global community reach its goal of net-zero carbon emissions. The center examines the accelerating energy transition and collaborates with industrial leaders to reform the world’s energy systems. The center is part of “Fast Forward: MIT’s Climate Action Plan for the Decade,” MIT’s multi-pronged effort announced last year to address the climate crisis.

The Future Energy Systems Center investigates the emerging technology, policy, demographics, and economics reshaping the landscape of energy supply and demand. The center conducts integrative analysis of the entire energy system — a holistic approach essential to understanding the cross-sectorial impact of the energy transition.


Thrilled to share that my article Information Privacy & the Inference Economy is forthcoming in @NwULRev . Draft on SSRN

Twitter, Alicia Solow-Niederman, SSRN


from

Information privacy is in trouble. Contemporary information privacy protections emphasize individuals’ control over their own personal information. But machine learning, the leading form of artificial intelligence, facilitates an inference economy that strains this protective approach past its breaking point. Machine learning provides pathways to use data and make probabilistic predictions—inferences—that are inadequately addressed by the current regime. For one, seemingly innocuous or irrelevant data can generate machine learning insights, making it impossible for an individual to anticipate what kinds of data warrant protection. Moreover, it is possible to aggregate myriad individuals’ data within machine learning models, identify patterns, and then apply the patterns to make inferences about other people who may or may not be part of the original data set. The inferential pathways created by such models shift away from “your” data, and towards a new category of “information that might be about you.” And because our law assumes that privacy is about personal, identifiable information, we miss the privacy interests implicated when aggregated data that is neither personal nor identifiable can be used to make inferences about you, me, and others.

This Article contends that accounting for the power and peril of inferences requires reframing information privacy governance as a network of organizational relationships to manage—not merely a set of data flows to constrain.


AI chip startup Ceremorphic comes out of stealth mode

ZDNet, Tiernan Ray


from

“It’s counterintuitive today, but higher performance is lower power, said Venkat Mattela, founder and CEO of the company, in an interview with ZDNet via Zoom.

Mattela believes that numerous patents on low-power operation will enable his company’s chip to produce the same accuracy on signature tasks of machine learning with much less computing effort.

“What I’m trying to do is not just building a semiconductor chip but also the math and the algorithms to reduce the workload,” he said. “If a workload takes a hundred operations, I want to bring it down to fifty operations, and if fifty operations cost less energy than a hundred, I want to say mine is a higher-performance system.”


How to Fix Big Tech’s Diversity Problem – Tech companies around the world could easily recruit more women and minorities if only they knew where to look.

Foreign Policy, Viewpoint, Bhaskar Chakravorti


from

After years of big promises and little change, Silicon Valley experienced a tiny breakthrough in raising diversity among its workforce, where women, Black, and Hispanic workers have long been underrepresented. On Jan. 12, Twitter said that it had boosted the proportion of Black employees at its U.S. locations to 9.4 percent from 6.9 percent in only one year and the share of Hispanic workers to 8.0 percent from 5.5 percent. Even if the company hasn’t revealed the seniority levels and functional areas where the hiring took place, the numbers attest to substantial changes, especially considering the lack of progress on diversity at other tech companies. How did it pull it off? Can others do the same? And can Twitter do even better? The answer is yes to all.

The reason Twitter could raise these numbers so quickly was summarized by the company’s vice president of inclusion, diversity, equity, and accessibility, James Loduca. Because of Twitter’s switch to allowing flexible work from anywhere, Loduca said, “We were able to hire folks in markets that we know have high populations of Black talent, markets that we know have high populations of Latinx talent.” Black candidates were more than twice as likely to accept a job offer from Twitter compared with the previous year, while Hispanic candidates were five times as likely.

If this strategy were used more widely, Twitter’s diversity gains could be copied elsewhere.


Stamford pulls back on plan to help universities expand in city

Stamford Advocate, Veronica Del Valle


from

A month after introducing a proposal that would create additional development parameters for universities looking to expand their footprint in Stamford, Mayor Caroline Simmons’ administration is pulling back on its plans.

City zoning officials recently announced Simmons’s office would “reassess” plans to create incentives and requirements for university-related development projects in Stamford.

“The Administration made a decision to revisit this item once our Director of Economic Development is in place,” chief of staff Bridget Fox said in a statement. “Creating a University and Research Overlay District is important for future economic growth in Stamford and we look forward to resubmitting the proposal in the coming months.”


Toward Ethical and Equitable AI in Higher Education

Inside Higher Ed, Beyone Transfer, Dan Knox and Zach Pardos


from

Machine learning–based grade prediction has been among the first applications of AI to be adopted in higher education. It has most often been used in early-warning detection systems to flag students for intervention if they are predicted to be in danger of failing a course, and it is starting to see use as part of degree pathway advising efforts. But how fair are these models with respect to the underserved students these interventions are primarily designed to support? A quickly emerging research field within AI is endeavoring to address these types of questions, with education posing particularly nuanced challenges and trade-offs with respect to fairness and equity.

Generally, machine learning algorithms are most accurate in predicting that which they have seen the most of in the past. Consequently, with grade prediction, they will be more accurate at predicting the groups of students who produce the most common grade. When the most common grade is high, this will lead to perpetuating inequity, where the students scoring lower will be worst served by the algorithms intended to help them. This was observed in a recently published study out of the University of California, Berkeley, evaluating predictions of millions of course grades at the large public university. Having the model give equal attention to all grades led to better results among underserved groups and more equal performance across groups, though at the expense of overall accuracy.


Dan Jurafsky receives 2022 Atkinson Prize in Psychological and Cognitive Sciences

Stanford University, School of Humanities and Sciences


from

Stanford professor of linguistics and of computer science receives award for his groundbreaking contributions to computational linguistics and the sociology of language with significant applications to machine learning, artificial intelligence, and social justice. [$100,000 prize]


RTI, UNC researchers develop smartphone app to fight neglected tropical disease

WRAL TechWire, Jason Parker


from

Researchers from RTI International and the University of North Carolina Chapel Hill released a new application that uses machine learning technology to detect trachomatous trichiasis.

Trachomatous trichiasis, according to RTI, is the painful end stage of the neglected tropical disease (NTD), trachoma, that can lead to blindness if not promptly diagnosed and treated with surgery.

January 30 marks World NTD Day, which Rebecca Flueckiger, associate director of Monitoring, Evaluation, Research, Learning and Adapting (MERLA) at RTI, told WRAL TechWire is a day that is “intended to raise awareness on neglected tropical diseases and the progress to eliminate them globally.”


This AI Could Be Robocallers’ Kryptonite

IEEE Spectrum, Rahul Rao


from

If it’s up to apps to cull nuisance calls, then it’s fortunate that they’re getting smarter. Some are now employing AI. One machine-learning-based system was created by a group including Mustaque Ahamad, a computer scientist at Georgia Tech, and Sharbani Pandit, then a graduate student there.

Ahamad describes the system as a “virtual assistant” that screens callers with a few questions, like “whom do you want to speak to?” or “what is the weather like where you are calling from?” By judging the answers—or details like whether a caller interrupts the questions—the natural language processing system can make an educated guess as to whether a call is legitimate (or wanted).


“It’s a bloodbath”: U.S. companies are pillaging Latin America’s tech talent

Rest of the World, Vittoria Elliott


from

It took Andrea Campos, the Mexico City–based founder of two-year-old mental health app Yana, six long months to find a senior front-end developer. After launching her app in the early days of the pandemic, Yana’s usership ballooned from just a few thousand users in Mexico to over 5 million across twelve countries. Campos, who said her company has raised $2.5 million in 2021 to scale the app and expand her team, was looking to bring on someone with the experience and skills to guide projects.

One month after his first day at Yana, the long-sought after developer told Campos he was leaving.

“An American company was offering him $15,000 per month to work for them,” Campos told Rest of World. “We cannot compete with that.”

Stories like Yana’s have become all too common across Latin America, where, according to every source who spoke to Rest of World and compiled salary data, demand for tech talent is skyrocketing, but supply remains relatively scarce, fueling fierce competition between startups, established tech companies, and outsourcing giants for qualified workers.


Everyone in New York Is Betting on Sports

New York Magazine, Intelligencer, Matt Stein


from

In the first ten days, 1.2 million accounts were created in New York with almost 90 percent of them new to sports betting, according to a geolocation firm that ensures gamblers are actually in the state they say they are. These bettors are pulling in staggering numbers for sports books: In its first three weeks, the total number of bets taken, known as the “handle,” has exceeded $1.175 billion, according to data from the New York State Gaming Commission.

Barry Sample, the chair of the gaming commission, expected such an impressive level of demand. When asked by email if the eight-figure haul was a surprise, he replied with the confidence of a house expert: “No.” The state’s monthly haul, which has been amplified by the insane early rounds of the NFL playoffs, is expected to exceed New Jersey this month to become the largest mobile sports-betting market in the country.


Let’s all take a moment to congratulate Juliana Freire (@jfreirenet ) and Yann LeCun (@ylecun ) for being elected as AAAS Fellows!

Twitter, NYU Center for Data Science


from


Today, @mikeitzkowitz published his Economic Mobility Rankings. Tonight you can see and interact with them.

Twitter, Jon Boeckenstedt, Michael Itzkowitz


from


Demystifying machine-learning systems – A new method automatically describes, in natural language, what the individual components of a neural network do.

MIT News


from

Neural networks are sometimes called black boxes because, despite the fact that they can outperform humans on certain tasks, even the researchers who design them often don’t understand how or why they work so well. But if a neural network is used outside the lab, perhaps to classify medical images that could help diagnose heart conditions, knowing how the model works helps researchers predict how it will behave in practice.

MIT researchers have now developed a method that sheds some light on the inner workings of black box neural networks. Modeled off the human brain, neural networks are arranged into layers of interconnected nodes, or “neurons,” that process data. The new system can automatically produce descriptions of those individual neurons, generated in English or another natural language.


Social-media platforms failing to tackle abuse of scientists

Nature, News, Brian Owens


from

Social-media sites such as Facebook and Twitter are not doing enough to tackle online abuse and disinformation targeted at scientists, suggests a study by international campaign group Avaaz.

The analysis, published on 19 January, looked at disinformation posted about three high-profile scientists. It found that although all of the posts had been debunked by fact-checkers, online platforms had taken no action to address half of them.

“Two years into the pandemic, even though they have made important policy changes, the platforms, and Facebook in particular, are still failing to take significant action,” says Luca Nicotra, a campaign director for Avaaz who is based in Madrid.


University of Minnesota student helping lead national ‘hackathon’ for female coders

StarTribune.com, Neal St. Anthony


from

Governess Simpson is a national leader in the technology field — and she hasn’t yet graduated from University of Minnesota.

Simpson, 22, besides studying industrial engineering and computer science, is one of the leaders of “Rewrite the Code Black Wings,” a 1,300-strong association of Black technology students from 215 universities. The group is part of the larger “Rewriting the Code,” a group of 14,000 women from 800 universities globally.

And Simpson, for the second year, also is a top planner for the upcoming “Black Wings Hacks” hackathon, which expects to attract more than 300 female college students. The virtual event will host teams that work on projects with mentors from sponsoring companies that showcase skills.


Events



Holding AI Accountable: Who Gets to Tell the Story?

Pulitzer Center


from

Join the Pulitzer Center, Northwestern University, and NU’s Medill School of Journalism, Media, Integrated Marketing Communications for a virtual conversation on AI accountability reporting on Thursday, February 24, at 12:00 pm EST (11:00 am CST).

SPONSORED CONTENT

Assets  




The eScience Institute’s Data Science for Social Good program is now accepting applications for student fellows and project leads for the 2021 summer session. Fellows will work with academic researchers, data scientists and public stakeholder groups on data-intensive research projects that will leverage data science approaches to address societal challenges in areas such as public policy, environmental impacts and more. Student applications due 2/15 – learn more and apply here. DSSG is also soliciting project proposals from academic researchers, public agencies, nonprofit entities and industry who are looking for an opportunity to work closely with data science professionals and students on focused, collaborative projects to make better use of their data. Proposal submissions are due 2/22.

 


Tools & Resources



When (not) to ban laptops from classrooms

Character Lab, Take Note blog, Daniel Oppenheimer


from

In my research, we showed students TED Talks and had them take notes either on laptops or with a pen and paper. Because most people can type faster than they can write, students using laptops transcribed the talks nearly verbatim. But students who took handwritten notes could not, so they took notes in their own words. Writing by hand required students to understand, synthesize, and summarize the content. Deeper processing of the information, in turn, led to improved learning. As a result, students in the longhand condition scored higher on tests.

Teachers around the country started banning laptops in their classes. But what actually interfered with learning was mindless transcription of a lecture. Laptops merely enabled that by allowing students to take notes more quickly.


Introducing PReF: Preprint Review Features

ASAPbio News


from

Preprint reviews hold the potential to build trust in preprints and drive innovation in peer review. However, the variety of platforms available to contribute comments and reviews on preprints means that it can be difficult for readers to gain a clear picture of the process that led to the reviews linked to a particular preprint.

To address this, ASAPbio organized a working group to develop a set of features that could describe preprint review processes in a way that is simple to implement. We are proud to share Preprint Review Features (PReF) in an OSF Preprint. PReF consists of 8 key-value pairs, describing the key elements of preprint review. The white paper includes detailed definitions for each feature, an implementation guide, and an overview of how the characteristics of active preprint review projects map to PReF. We also developed a set of graphic icons (below) that we encourage the preprint review community to reuse alongside PReF.


Careers


Tenured and tenure track faculty positions

They’re letting me hire someone else to help teach stats/data science @EmoryOxford



Twitter, Zachary Binney

Leave a Comment

Your email address will not be published.