Data Science newsletter – December 15, 2020

GROUP CURATION: N/A

Prices and wait times on @lyft
app were randomized across 14 million sessions in a dozen large US markets

Coupling price & time randomization with Gary Becker’s seminal time allocation model, we can estimate the VOT for the average person, and across many different people, choice circumstances, & market conditions.

Upshot: the VOT can vary over economic circumstances!

What distinguishes the AI of today from the AI of yesterday and other digital technologies that we’ve become accustomed to over the last several decades? Machine learning. It is currently the dominant technique in AI that is causing much of the hubbub over AI and its ethical implications. Rather than pre-programming an expert system that is told what to do by its code, machine learning relies on algorithms that learn from data over time in order to do things like classify objects or make predictions.

Object of Analysis

Given this definition of AI, what exactly is the object of study for the sociologist? The sociologist needs to examine AI from its most expansive conceptualization possible to trace its full reach into and across societies. I think of it as originating with the AI model, the difficult-to-get-to, blinding core of the sun. (Many think of this as the “black box,” but I like a slightly sunnier metaphor.) From there, the model emanates like sun rays all the way to the most macro-level social structures of human society, impacting everything it touches. (On second thought, given the countless negative “emanations” of AI perhaps this metaphor is too sunny after all.) The point is that while the algorithm’s code is relatively bounded, its effects radiate far and wide. It is hard for most of us to make the mental leap from some lines of code to the destruction of Western democracy. The sociologist can help us connect the dots across the necessary levels of analysis.

Microsoft unveiled an initiative to use artificial intelligence to detect government corruption, calling it “an urgent global issue that can and must be solved.”

The new Microsoft Anti-Corruption Technology and Solutions initiative “will leverage the company’s investments in cloud computing, data visualization, AI, machine learning, and other emerging technologies to enhance transparency and to detect and deter corruption” over the next decade, said Dev Stahlkopf, Microsoft general counsel, in a post announcing the plan.

Microsoft made the announcement in conjunction with the United Nations’ International Anti-Corruption Day. Stahlkopf described the initiative as increasingly important given the events of the past year.

Who do you think performs your medical laboratory tests for COVID-19 or any other test? If you answered “my doctor” or “my nurse” or a robot, you would be completely wrong.

To put it bluntly, your life is in the hands of medical laboratory professionals. We perform an estimated 13 billion laboratory tests in the United States each year. That means that laboratory testing is the single highest-volume medical activity in the lives of Americans.

Why should you care? Those 13 billion tests help drive approximately two-thirds of all medical decisions made by your doctor and other health care professionals from cradle to grave. There are only 337,800 practicing medical laboratory professionals for a population of just over 330 million people in the U.S.

Alberta’s chief medical officer is defending the decision to allow the world junior hockey championship to be played, even while cases mount across Canada during a second wave of the COVID-19 pandemic.

Dr. Deena Hinshaw said Monday she feels the tournament can be safe.

“My team and I have worked with the organizers of the world junior hockey tournament to make sure their protocols are going to protect public safety, that they will be adequate to make sure the tournament can proceed in a safe way,” Hinshaw said.

Ecologists love to study blue tits. The birds readily nest in boxes in the wild and have striking plumage that seems ideal for testing ideas about the evolutionary point of the ornamentation. Dozens of studies have reported that male coloring is substantially different from that of females, that females choose mates based on differences in that coloring, and that male plumage is a signal of mate quality.

But Tim Parker, an ecologist at Whitman College, wasn’t so sure. In a 2013 meta-analysis of 48 studies on blue tit plumage, Parker found many researchers had cherry-picked the strongest findings from data they had sliced and diced. They had worked backward from results to form hypotheses that fit the data. And reams of boring, negative results were missing from the published picture. There was no reason to think these problems were limited to blue tits, Parker says: “I just became convinced that there was a lot of unreliable stuff out there.”

Parker soon found an ally in Shinichi Nakagawa, an ecologist at the University of New South Wales with similar concerns. “It’s an existential crisis for us,” Nakagawa says. The two began to publish on the issue and gathered more collaborators. That has culminated with the launch last week of the Society for Open, Reliable, and Transparent Ecology and Evolutionary Biology (SORTEE).

The last thing UW-Madison junior Paige Leistner needed this semester was additional stress.

On top of a rocky transition to online classes, a fear of contracting COVID-19 and limited interactions with friends, Leistner also has to deal with another unsettling source of anxiety — the anti-cheating computer software that monitors her every move during an exam.

Welcome to the world of Honorlock, the online proctoring software more than 300 universities and colleges, including UW-Madison and Madison Area Technical College, have turned to for assistance in testing students’ knowledge during the pandemic.

Artificial Intelligence, an enigmatic term for technologies that make gadgets and software “smart,” is expected to become a bigger part of our lives thanks to advances in computing power, data storage and high-speed networks such as 5G.

San Diego is in a strong position to benefit from the expansion of artificial intelligence, according to a study “Measuring the Future: AI and San Diego’s Economy” released last week from the San Diego Regional Economic Development Corp.

While the study did not pinpoint a specific number of artificial intelligence jobs in the region, it did highlight industries with above-average employment in AI fields.

They include telecommunications, information technology, software and transportation.

Choosing the right analysis model and tool to do the proper job for high throughput data analysis remains a great challenge. So the FDA invited a team of researchers at the Georgia Institute of Technology to conduct a comprehensive investigation of RNA-seq data analysis pipelines for gene expression estimation to recommend best practices.

“No common standard for selecting high throughput RNA-seq data analysis tools has been established yet. This has been a huge challenge for studying hundreds of tools that form tens of thousands of analysis pipelines,” noted May Dongmei Wang, a professor in the Wallace H. Coulter Department of Biomedical Engineering at Georgia Tech and Emory University who led the investigation.

Huawei has been running a series of roundtables talking about the enormous problems that need to be addressed to make the world a better and safer place. This week the topic was on “Unleashing Innovation Through Collaboration,” and the framework was the COVID-19 response.

On the panel were Susan Athey, a professor and Stanford graduate who wrote The Economics of Technology, Antony Walker, Deputy Chief Executive Officer for TechUK, Martina Szabo, Business Development and Strategy Lead for the World Economic Forum, and Andy Purdy CSO for Huawei Technologies USA. The panel was moderated by Simon Cox, Emerging Markets Editor for the Economist (and he did an excellent job moderating).

Early in the year, when it first became clear we faced a pandemic, technology companies worldwide stepped up and pledged resources that should have been able to find the best ways to balance economic impact and safety. But while the response was faster than the 1918 Flu Virus response, it wasn’t that much faster, and tons of mistakes should have been avoidable given we have massive modeling capability. At the heart of the problem wasn’t the lack of data; it was the inability to get to that data and analyze it in a timely basis.

In 2013, the University of Texas at Austin’s computer science department began using a machine-learning system called GRADE to help make decisions about who gets into its Ph.D. program — and who doesn’t. This year, the department abandoned it.

Before the announcement, which the department released in the form of a tweet reply, few had even heard of the program. Now, its critics — concerned about diversity, equity and fairness in admissions — say it should never have been used in the first place.

“Humans code these systems. Humans are encoding their own biases into these algorithms,” said Yasmeen Musthafa, a Ph.D. student in plasma physics at the University of California, Irvine, who rang alarm bells about the system on Twitter. “What would UT Austin CS department have looked like without GRADE? We’ll never know.”

The world’s most closely watched AI conference didn’t have its usual scenic backdrop of Vancouver or Barcelona, but it was a nicely organized six-day virtual affair that delivered on all the usual fascinating talks and presentations.

The Board of Visitors Buildings and Grounds Committee on Friday approved the schematic design of the proposed home of the School of Data Science in the Emmet-Ivy Corridor.

The School of Data Science building is scheduled to be the first academic building constructed on the 14.5-acre Emmet-Ivy parcel. As designed, the building will face Emmet Street, with an amphitheater and a retention pond between the building façade and the street.

The four-story, 61,000 square-foot facility will include four “adaptive classrooms” with technology to enhance the learning process, and faculty offices, as well as meeting and research areas that will support the School of Data Science.

Creating healthy habits, like increasing physical activity and improving eating habits, can be difficult to maintain long term, especially without accountability. Research from the Center for Weight, Eating and Lifestyle Science (WELL Center) in the College of Arts and Sciences at Drexel University suggests that health counselors having access to self-monitored health data would improve a person’s weight loss maintenance.

The results of several ambitious studies testing wearables as early predictors of for Covid-19 are in — and they suggest that data from devices including Apple Watches, Fitbits, and Oura smart rings may be useful for flagging some infections in people before they even feel ill.

Recently published research from ongoing efforts at three high-profile institutions in the Golden State — the University of California in San Francisco, Stanford University, and Scripps Research Translational Institute in San Diego — indicate that wearables can detect a bump in heart rate or temperature, the most consistent signs that the body is mounting a response to an external threat before symptoms appear. Feeding those data to algorithms that crunch large amounts of information provides a sort of traffic map for the spread of the virus — and could prove a useful tool in the pandemic response in the months to come.

While the Scripps and Stanford studies are device-agnostic — they accept data from people with Apple Watches, Fitbits, and Garmins — the UCSF study is a partnership with smart ring maker Oura and analyzes data from those devices only.

“The R Consortium’s COVID-19 Working Group is providing a new home for the COVID-19 Data Hub Project. The goal of the COVID-19 Data Hub is to provide the worldwide research community with a unified dataset by collecting worldwide fine-grained case data, merged with external variables helpful for a better understanding of COVID-19.”

“Support for spatial data and spatial indexing is one of the most requested features in the history of CockroachDB. The first issue requesting spatial data in CockroachDB was opened in October 2017, and closed on November 12, 2020 with the release of spatial data storage in CockroachDB 20.2.”

“NIST researchers pulled together a decade’s worth of fire experiments into a robust online resource for everyone to access. Explore the Fire Calorimetry Database”