Data Science newsletter – December 14, 2021

The central question that sports scientists are grappling with these days is this: What the heck are we going to do with all this data? In endurance sports, we’ve progressed from heart rate monitors and GPS watches to sophisticated biomechanical analysis, internal oxygen levels, and continuous glucose measurements, all displayed on your wrist then automatically downloaded to your computer. Team sports have undergone a similar tech revolution. The resulting data is fascinating and abundant, but is it actually useful?

A new paper in the International Journal of Sports Physiology and Performance tackles this question and presents an interesting framework for thinking about it, derived from the business analytics literature. The paper comes from Kobe Houtmeyers and Arne Jaspers of KU Leuven in Belgium, along with Pedro Figueiredo of the Portuguese Football Federation’s Portugal Football School.

A team led by scientists at the London-based artificial-intelligence company DeepMind has developed a machine-learning model that suggests a molecule’s characteristics by predicting the distribution of electrons within it. The approach, described in the 10 December issue of Science1, can calculate the properties of some molecules more accurately than existing techniques.

“To make it as accurate as they have done is a feat,” says Anatole von Lilienfeld, a materials scientist at the University of Vienna.

The paper is “a solid piece of work”, says Katarzyna Pernal, a computational chemist at Lodz University of Technology in Poland. But she adds that the machine-learning model has a long way to go before it can be useful for computational chemists.

It was a tally that shocked the experts: 38,680 deaths on U.S. roadways last year, the most since 2007, even though pandemic precautions had dramatically reduced driving.

“This was completely unprecedented,” said Ken Kolosh, a researcher at the nonprofit National Safety Council. “We didn’t know what was happening.”

One possibility was that stressed-out Americans were releasing their anxieties on the wide-open roads. He guessed that fatal accidents would decline in 2021 when traffic returned.

He was wrong.

Meat consumption is growing globally. But a handful of countries are bucking this trend, and their appetite for meat is in decline. A group of researchers argues that these nations—New Zealand, Canada, and Switzerland—have in fact reached ‘peak meat’, a point beyond which increasing income no longer tracks with increasing consumption of beef, chicken, mutton, and pork.

This could be an important discovery, because reducing meat consumption is recognized as a critical route to bringing down greenhouse gas emissions. In light of this, the group of Australian researchers were curious to see how consumption trends changed between 2000 and 2019, which covers a period of growing awareness about meat’s impact on the planet, but also a time when those environmental impacts intensified. They examined levels of meat consumption in 35 countries over this period, and combined this with information about gross domestic product (GDP), a measure of the size and health of a country’s economy.

Student enrollments in William & Mary’s data science courses (including summer courses) continue to rise. Currently there are 79 students pursuing the B.S. degree.

“Students were asking for our help building pieces of data science into self-designed majors that integrated their academic interests,” said Matthias Leu, associate professor of Biology current director of William & Mary’s Data Science program. “A core group of faculty started talking about how we might design a formal academic program, and the excitement just keeps building.”

To date, the faculty have designed and garnered formal approval from the Commonwealth of Virginia for a data science minor and a major leading to a B.S. degree. Next up: creating a discrete academic unit embedded with the Department of Computer Science. After that, focus will be on designing a second major leading to a B.A. degree.

The most popular college major in America is business administration. Often thought to lead to well-paying jobs, it attracts particularly high percentages of minority students. However, the report suggests that, at least at non-selective schools such as community colleges, business is not as profitable as it may seem. Business majors at non-selective schools earned less money on average than their classmates who focused on other subjects.

“It just came out looking awful,” said Deborah M. Weiss, director of the Workforce Science Project at the CLBE and one of the report’s authors. The earnings for business majors at more highly-selective schools were closer to average.

For students at non-selective schools looking for a more lucrative career, the report points to a different path.

“Nursing is an unbelievable major,” said Weiss.

Gaming has been at the center of the ML renaissance for the last few years. In many ways, the deep learning race was kicked off by AlphaGo performance again Go’s legend Le Sedol. However, most of the advancements in ML for gaming have been specialized in either perfect or imperfect information games but never both at the same time. Models that mastered chess and Go struggle with imperfect games such as Poker. Even models like AlphaZero, which learned to play multiple games simultaneously, were constrained to perfect information environments. The reason for this is based on the intrinsic dynamics of both types of game environments. Perfect information games, like Chess and Go, are a good fit for ML techniques relying on self-play learning and local min-max search of the gamespace, while imperfect information games like Poker rely on game-reasoning techniques. Earlier this week, DeepMind unveiled a new ML method that can change these constraints forever.

Player of Games (PoG) is the first ML model able to play perfect and imperfect information games at scale. Created by DeepMind, PoG combines self-play learning, search, and game-theoretic reasoning in a single model

As University of Miami students and alumni celebrate the hiring of a new football coach who has been touted as the program’s savior, not everyone is doing cartwheels on the Coral Gables campus.

Many UM professors and other faculty — forced to teach in person and accept compensation cuts during the COVID-19 pandemic — are fuming and say staff morale has plunged. Some members of the UM Faculty Senate view Mario Cristobal’s $80 million contract as only the latest sign of the administration’s disrespect for the academic staff at the private university.

Machines that understand and operate user interfaces (UIs) on behalf of users could offer many benefits. For example, a screen reader (e.g., VoiceOver and TalkBack) could facilitate access to UIs for blind and visually impaired users, and task automation agents (e.g., Siri Shortcuts and IFTTT) could allow users to automate repetitive or complex tasks with their devices more efficiently. These benefits are gated on how well these systems can understand an underlying app’s UI by reasoning about 1) the functionality present, 2) how its different components work together, and 3) how it can be operated to accomplish some goal. Many rely on the availability of UI metadata (e.g., the view hierarchy and the accessibility hierarchy), which provide some information about what elements are present and their properties. However, this metadata is often unavailable due to poor toolkit support and low developer awareness. To maximize their support of apps and when they are helpful to users, these systems can benefit from understanding UIs solely from visual appearance.

Recent efforts have focused on predicting the presence of an app’s on-screen elements and semantic regions solely from its visual appearance. These have enabled many useful applications: such as allowing assistive technology to work with inaccessible apps and example-based search for UI designers. However, they constitute only a surface-level understanding of UIs, as they primarily focus on extracting what elements are on a screen and where they appear spatially. To further advance the UI understanding capabilities of machines and perform more valuable tasks, we focus on modeling the higher-level relationships by predicting UI structure.

Amid the wall décor in Karen Levy’s Gates Hall office—across from an assortment of trucker memorabilia and a signed photo of late Supreme Court Justice Ruth Bader Ginsburg ’54—is a poster from the classic thriller The Conversation. A Best Picture nominee written and directed by Francis Ford Coppola, the 1974 film stars Gene Hackman as a surveillance expert tasked with bugging an “unrecordable” event: two people talking as they walk through a busy, noisy city square.

It’s apt artwork for Levy, an assistant professor of information science whose research explores the often fraught intersection of technology and privacy. Levy, who also has an appointment in the Law School, had already earned a JD from Indiana University and clerked for a federal judge when she decided to pursue a PhD in sociology from Princeton—training that gives her a novel perspective as she explores a wide variety of topics, from the use of webcams in nursing homes to the ways in which retailers track customers to the role of technology in intimate partner abuse.

Tracking on 18 wheels

But first, there were the truckers. While the current shortage of long-haul drivers—and its role in pandemic-related supply chain issues—has made headlines in recent months, Levy began studying the industry a decade ago. The subject of her doctoral thesis (and the reason why she has a trucker patch and belt buckle on her office wall) was how surveillance technology has impacted trucking.

Weighing the likelihood of an outcome, or the nature of risk, is decidedly difficult — especially when those decisions involve complex systems. How should a policymaker compare the health risks of concentrating pollution in a local environment versus diffusing the pollution more globally? How should societies allocate scarce resources like hunting tags — should benefits be distributed by uniform lotteries, or guaranteed to some subset of people? These and other fraught queries are the concern of a group of Cornell Ann S. Bowers College of Computing and Information Science researchers who have directed their attention to algorithmic decision-making in the context of allocating harms and benefits — and now their findings have achieved professional acclaim.

The paper, “On Modeling Human Perceptions of Allocation Policies with Uncertain Outcomes,” which won the Exemplary Applied Modeling Track award during the 22nd Association for Computing Machinery Conference on Economics and Computation (ACM EC 21), combines mathematical and qualitative analysis to study situations in which society allocates harms or benefits that are uncertain in nature, and, in turn, proposes explanations for societal preferences that would otherwise be difficult to explain using standard models of cost-benefit analysis.

South Korea will soon roll out a pilot project to use artificial intelligence, facial recognition and thousands of CCTV cameras to track the movement of people infected with the coronavirus, despite concerns about the invasion of privacy.

The nationally funded project in Bucheon, one of the country’s most densely populated cities on the outskirts of Seoul, is due to become operational in January, a city official told Reuters.

The system uses an AI algorithms and facial recognition technology to analyse footage gathered by more than 10,820 CCTV cameras and track an infected person’s movements, anyone they had close contact with, and whether they were wearing a mask, according to a 110-page business plan from the city submitted to the Ministry of Science and ICT (Information and Communications Technology), and provided to Reuters by a parliamentary lawmaker critical of the project.

High-school juniors applying to Stanford University can stop studying for standardized tests. The school isn’t requiring them. Columbia University said last month it won’t require scores from students who are now sophomores or juniors, joining Cornell University and Amherst College. And the 280,000-student University of California system has declared no testing for all freshmen who may apply to its 10 campuses.The Covid-inspired movement that freed high-schoolers from the all-encompassing dominance of the SAT is rippling through higher education and likely to persist beyond the pandemic. “Left on our own, I think we’ll go the way of the University of California system and either be completely test-optional or as far as score-free for a long time to come,” said Jonathan Burdick, who oversees enrollment at Cornell, where three of the seven undergraduate colleges won’t accept scores.{“body”:”High-school juniors applying to Stanford University can stop studying for standardized tests. The school isn’t requiring them. Columbia University said last month it won’t require scores from students who are now sophomores or juniors, joining Cornell University and Amherst College. And the 280,000-student University of California system has declared no testing for all freshmen who may apply to its 10 campuses.

The Covid-inspired movement that freed high-schoolers from the all-encompassing dominance of the SAT is rippling through higher education and likely to persist beyond the pandemic. “Left on our own, I think we’ll go the way of the University of California system and either be completely test-optional or as far as score-free for a long time to come,” said Jonathan Burdick, who oversees enrollment at Cornell, where three of the seven undergraduate colleges won’t accept scores.

The pandemic reshaped admissions when it closed testing sites, prompting schools to use a wider variety of measures to decide who is admitted. The SAT and ACT, which assess math and verbal skills, have long been criticized as a barrier to diversity, because wealthier students can afford tutors and fancy consultants. Testing companies say they’re a neutral way to measure raw talent and reward hard work.

The more money people come across, the more things they can and tend to buy. More money on average means bigger houses, more expensive cars, and fancier restaurants. But what if you look at relative spending instead of total dollars?

For example, if a lower income group uses 9 percent of their total spending to pay a mortgage, does the higher income group also pay 9 percent? Or does additional income go to other spending categories?

It varies.

The charts below show how different income groups spend their money, based on data from the Bureau of Labor Statistics for 2020. Each chart represents a spending category. Each column represents an income group

Financial regulators are “three or four steps behind” technological innovation — and that gets dangerous,” argues Andrew W. Lo, finance professor at MIT Sloan School of Management and director of the MIT Laboratory for Financial Engineering, in an interview with ThinkAdvisor.

The “biggest challenge” for regulators is trying to keep pace with the years-long pick-up in tech innovation, says Lo, who is on FINRA’s Economic Advisory Committee.

Financial regulation is just one area of his current research.

For another, he is a principal investigator at the MIT Computer Science and Artificial Intelligence Laboratory.

On Friday, needing just one point against Ian Nepomniachtchi to defend his world champion status, Magnus Carlsen closed the match out with three games to spare, 7.5-3.5. He’s been the No 1 chess player in the world for a decade now and is in his eighth year as undisputed world champion.

“What happened here has never happened before in my career,” Nepomniachtchi said after Friday’s game, in a perhaps unintended homage to the new champion. “I’ve lost lots of stupid games, but not as many in such a short time.”

What if your boss was an algorithm? What would you do if your employer suddenly fired you or reduced your pay without telling you why? And without being willing to give you a reason when you ask for one?

This is not science fiction or some far-fetched reality. Millions of people worldwide are working in the gig economy sector for companies like Uber, Deliveroo, Bolt, Just Eat… And this could be the future of work for people working outside the gig economy, as surveillance technologies are creeping into the workplace – and the ‘work-from-home place’ in particular.

Who we are working with

To counter the surveillance that employers are subjecting workers to, and the power imbalance that workers face, we have partnered with Worker Info Exchange and App Drivers and Couriers Union, who have been working on these issues and fighting to protect rights of gig economy workers.

A global hacker “red team” and rewards for hunting algorithmic biases are just some of the recommendations from experts who argue that AI faces a “tech-lash” unless firm measures are taken to increase public trust.

The Artificial Intelligence industry should create a global community of hackers and “threat modellers” dedicated to stress-testing the harm potential of new AI products in order to earn the trust of governments and the public before it’s too late.

This is one of the recommendations made by an international team of risk and machine-learning experts, led by researchers at the University of Cambridge’s Centre for the Study of Existential Risk (CSER), who have authored a new “call to action” published in the journal Science.