Data Science newsletter – December 9, 2020

GROUP CURATION: N/A

But for all the claims that pandemics prove urban density is bad, recent studies are suggesting the opposite. Data collected from 284 Chinese cities by The World Bank found that urban density may not be as much of an enemy in the coronavirus fight than we thought. In fact, cities with very high population densities such as Shanghai, Beijing, and Shenzhen had far fewer confirmed cases per 10,000 people than cities with lower population densities.

Similarly, one study found no association between the population density of 36 world cities (as measured in people per square kilometre) and rates of Covid-19 cases or deaths. And a study of 913 US metropolitan counties found that density is not significantly related to higher infection rate of COVID-19. And this may have more to do with behaviour than available space.

You may be surprised to learn that of the trio of long-awaited coronavirus vaccines, the most promising, Moderna’s mRNA-1273, which reported a 94.5 percent efficacy rate on November 16, had been designed by January 13. This was just two days after the genetic sequence had been made public in an act of scientific and humanitarian generosity that resulted in China’s Yong-Zhen Zhang’s being temporarily forced out of his lab. In Massachusetts, the Moderna vaccine design took all of one weekend. It was completed before China had even acknowledged that the disease could be transmitted from human to human, more than a week before the first confirmed coronavirus case in the United States. By the time the first American death was announced a month later, the vaccine had already been manufactured and shipped to the National Institutes of Health for the beginning of its Phase I clinical trial. This is — as the country and the world are rightly celebrating — the fastest timeline of development in the history of vaccines. It also means that for the entire span of the pandemic in this country, which has already killed more than 250,000 Americans, we had the tools we needed to prevent it .

To be clear, I don’t want to suggest that Moderna should have been allowed to roll out its vaccine in February or even in May, when interim results from its Phase I trial demonstrated its basic safety.

Water joined gold, oil and other commodities traded on Wall Street, highlighting worries that the life-sustaining natural resource may become scarce across more of the world.

Farmers, hedge funds and municipalities alike are now able to hedge against — or bet on — future water availability in California, the biggest U.S. agriculture market and world’s fifth-largest economy. CME Group Inc.’s January 2021 contract, linked to California’s $1.1 billion spot water market, last traded Monday at 496 index points, equal to $496 per acre-foot.

The contracts, a first of their kind in the U.S., were announced in September as heat and wildfires ravaged the U.S. West Coast and as California was emerging from an eight-year drought.

… Smartwatches and other wearable devices are traditionally used to track fitness but scientists at Stanford University found they may have a higher purpose in fighting the coronavirus pandemic.

Years ago, Dr. Michael P. Snyder detected he had Lyme disease because his smartwatch alerted him to an unusual resting heart rate.

Now, he along with his team are using that same model for early coronavirus detection, even as far in advance as four days before any symptoms may appear.

“What’s powerful about them is they’re measuring you all the time, so when you get ill it turns out your heart rate jumps up usually for regular illness about a day and a half before symptoms appear but for COVID-19 it turns out the median time is four days before symptoms appear,” Dr. Michael P. Snyder said.

Conservation scientist Kim Williams-Guillén was wracking her brain to come up with a way to save endangered sea turtles from egg poachers when she had an “aha” moment: If she placed a fake egg containing a GPS tracker in the reptiles’ nests, she might be able to track the thieves.

The idea won her the 2015 Wildlife Crime Tech Challenge—and a $10,000 prize. Now, Williams-Guillén, a conservation scientist at the environmental nonprofit Paso Pacífico, and a multinational team of colleagues have not only made the device—dubbed the InvestEGGator—but have also published the results of their first field test. Of 101 decoy eggs, five were able to track the routes of poachers up to hundreds of kilometers away. The “amazing” approach could one day help identify and stop high-level traffickers in the trade chain, says Héctor Barrios-Garrido, a conservation biologist from the University of Zulia, Maracaibo, who was not involved with the study.

Contact tracing, along with testing and isolating people who are infected, is considered crucial for controlling the coronavirus’s spread until a vaccine becomes widely available.

Health department contact tracers try to notify and interview anyone who was within 6 feet of an infected person for a total of 15 minutes or more. Studies show that, to be effective, that notification needs to happen within a few days of the person’s symptoms appearing. But case numbers have gotten so high, it’s become impossible for many states and counties to keep up. Some, including in Alaska, Missouri and Wisconsin, have urged residents to start notifying contacts on their own.
The time spent waiting in line for COVID-19 testing could be used to start your own contact tracing.

We created CogTracer as a free online tool to help people start tracing those contacts using best-practice interviewing methods.

George Serafeim wants to revolutionize the way businesses calculate their success.

Profit and loss aren’t enough, says the Harvard Business School professor. Serafeim aims to do what no one has done before: Put a dollar value on the impact of products and operations on people and the planet, then add or subtract it from companies’ bottom lines.

Intel Corp. provides an example of both. Serafeim and his five-person team credited $6.9 billion to the chipmaker in 2018 for paying its employees well and for boosting local economies where it has offices. But they deducted $3.1 billion for what they said was a shortage of women employees, the difficulty of career advancement and not enough attention paid to workers’ health.

“Without monetizing impacts, we’re left with the illusion that businesses have no impact,” Serafeim said. Companies that show big profits can have enormous negative effects on society, he said. “They’re just cheating because they’re operating in a context that doesn’t price all those impacts.”

Although Muslim Pro announced it would stop selling data to X-Mode, the awful truth remains: far too many companies that collect geolocation data can make a quick buck by selling that information, and the federal government is a regular buyer.

This violates the First and Fourth Amendments. In the current location data marketplace, if your phone and apps know where you are, then the government could, too. But the Supreme Court has decided that our detailed location data is so revealing about our activities and associations that law enforcement must get a warrant in order to acquire it. Government purchase of location data also threatens to chill people’s willingness to participate in protests in public places, associate with who they want, or practice their religion. History and legal precedent teach us that when the government indiscriminately collects records of First Amendment activities, it can lead to retaliation or further surveillance.

Unfortunately, this problem goes well beyond X-Mode. Other data brokers that sell app-derived location data to the federal government include Anomaly Six, Locate X, and Venntel.

Colleges and universities are much more facile at outside innovation than inside innovation. By that, I mean that they have created cultures and rewards for moving ideas and inventions outside of the institution through technology transfer or student education but are not as adept at achieving change within.

For fifteen years, while serving as vice president of research at the University of Southern California (USC), I witnessed—and encouraged—a transformation in research practices. Under the umbrella of the “Creativity and Collaboration in the Academy” initiative, USC explored and experimented with how higher education institutions might evolve to a research culture that, aligned with information technology, could achieve innovation inside the institution so as to be more impactful on the outside.

The technology behind WA Notify has been vetted by security & privacy experts across the world…I plan to add WA Notify to my phone and I will encourage my friends & family to use it.” 1/2

Want to learn more about the privacy-preserving principles underpinning WA Notify? Check out our team’s work on mobile contact tracing with @MSFTResearch
from April, part of #UWAllen’s rapid response to the #COVID19 pandemic: https://news.cs.washington.edu/2020/04/08/privacy-and-the-pandemic-uw-researchers-present-a-pact-for-using-technology-to-fight-the-spread-of-covid-19/ #UWserves #UWinnovates 2/2

From the email it seems that Google wanted her to tone down the critical nature of their work! The company’s internal filter wanted them to add that latest AI models are getting better than previous models both environmentally and in fairness as well! 2/n

In late October 2020, venture capital firm Wing conducted a survey, “Chief Data Scientist Survey,” of 320 of the senior-most data scientists at both global corporations and venture-backed startups, in advance of its annual Wing Data Science Summit.

Among the survey’s findings:

Explainability is data scientists’ #1 challenge with respect to models currently, at 46 percent.

Data labeling (29 percent), model deployment (28 percent) and data quality checks (25 percent) round out the top 4.

A Fortune 10 data science leader said that business users tell his team, “You’re doing this magic. I don’t trust it yet. That’s why I like to use the rules we have used forever, even if your models show they are performing at only 30-40 percent.”

When asked what technology developments will be most important for data science and machine learning teams over the next 12 months, again Explainability was the #1 choice.

While public health officials suspect off-field interactions may be contributing to community spread, there’s little hard data. In most areas, there have been few to no documented outbreaks, much less superspreader events.

Ice hockey is an anomaly. Scientists are studying hockey-related outbreaks hoping to find clues about the ideal conditions in which the coronavirus thrives – and how to stop it. Experts worry that ice rinks may trap the virus in an area that, by design, restricts airflow, temperature and humidity.

The hockey-related cases have been especially striking, epidemiologists have said, because clubs followed Centers for Disease Control and Prevention limits on gathering size and had numerous social distancing measures in place. In retrospect, one mistake by some clubs was that until recently masks had only been required on ice for the two players doing the initial faceoff for the puck – although many players wore clear face shields, which theoretically should have a similar effect.

“We’re watching hockey very carefully because it’s the first major sport that’s been played indoors predominantly and also during the winter months,” said Ryan Demmer, an epidemiologist at the University of Minnesota’s School of Public Health.

“In ‘A New Type of Leader: Harnessing Humor as a Secret Weapon,’ Professor Jennifer Aaker and Lecturer Naomi Bagdonas draw from their experience in teaching the GSB course ‘Humor: Serious Business’ and their latest book Humor, Seriously: Why Humor Is a Secret Weapon in Business and Life, in discussing the power of humor and how you can harness it as a leader to disarm, encourage bold actions and give other people courage to do the same.”

These are my favorite resources for deep learning news, which I’ve accumulated over the past 4 years working as a Machine Learning Engineer. I love them, because they save me time and effort. There are simply not enough hours in a week to manually scour the ArXiv for research papers. These sites can help to filter that down, and show you only the high-quality content you need.

I also publish a weekly update on deep learning, which contains research papers, blog posts, and Github repos I liked from the past week. Follow me to get a condensed, curated list of deep learning news each week!