Data Science newsletter – October 29, 2020

Newsletter features journalism, research papers and tools/software for October 29, 2020

GROUP CURATION: N/A

 

Companies Are Rushing to Use AI—but Few See a Payoff

WIRED, Business, Will Knight


from

At some DHL shipping centers, artificial intelligence now helps employees make sure pallets will load safely into cargo planes. A computer vision system captures each pallet, and an algorithm judges whether it can be stacked with other pallets or may be too awkward to fit on the next flight.

DHL is one of a growing number of companies using AI. Besides the pallet scanning system, AI helps route deliveries, control robots that ferry packages around warehouses, and control an experimental robot arm that picks and sorts parcels. DHL is also among a small minority of companies using AI—just 11 percent—that say they’ve reaped a significant return on investment from using the technology, according to a new report.

The report, from Boston Consulting Group and MIT Sloan Management Review, is one of the first to explore whether companies are benefiting from AI. Its sobering finding offers a dose of realism amid recent AI hype. The report also offers some clues as to why some companies are profiting from AI and others appear to be pouring money down the drain.


DC Rainmaker State of Sports Tech 2020 Keynote

DC Rainmaker blog, Ray Maker


from

… the focus of the presentation begins with some of the overarching trends in the wearables and sports technology space, and then gets successively deeper into the weeds as we go along, talking about areas such as indoor training (and the rise of anti-competitive practices), or power meters and even running power. I also talk about the impact of COVID-19 on sports tech as well.


How Sight—Not Taste, Smell, or Touch—Became the Sense of the Supermarket

Behavioral Scientist, Ai Hisano


from

I discovered people’s sensory experience in food shopping and eating changed as a product of complex interplay between advances in science and business, how those advances were reflected by our attitude toward nature, and the introduction of new political and economic powers. Over time, the forces of industrialization and standardization pushed for a bright, uniform look of food available year-round, exposing people to new visual and taste experiences. But this came at a cost to the other senses and diminished more localized relationships with food.

Eye appeal has become buy appeal. But why and how? A history of how they became connected provides insight into the transformation of people’s relationships with food, nature, and society.


How obesity could create problems for a COVID vaccine

Nature, News Feature, Heidi Ledford


from

Researchers fear that vaccines might not be as effective in people who are obese, a population already highly vulnerable to COVID-19.


The invader – How the new coronavirus penetrates, exploits and kills cells, and how an army of scientists aims to destroy it

Stanford Medicine magazine, Bruce Goldman


from

“Know your enemy,” Sun Tzu, the great sage of war, wrote some 2,500 years ago. Today, as COVID-19 spreads around the globe, the greatest army of medical scientists ever assembled is bent on learning all it can, as fast as it can, about SARS-CoV-2, the virus behind the pandemic.

Here’s a primer on viruses in general and SARS-CoV-2 in particular. As researchers learn more and more about the novel coronavirus that causes COVID-19, this knowledge — gathered through unmatched levels of scientific cooperation — is being turned against the virus in real time.


2U, Netflix Partner With Norfolk State University to Launch Technology Boot Camps

Diverse: Issues In Higher Education, Sarah Wood


from

The 16-week boot camps will focus on the areas of applied data science, advanced Java programming and UX/UI design. NSU faculty and guest technology experts will teach participants industry skills while also applying the curriculum to real world challenges within the sector.

“You may study these kinds of things in school but to have a chance to go through a four-month boot camp where you are basically grappling with some of the same real world challenges that Netflix does,” said David Sutphen, chief strategy and engagement officer at 2U. “That is the other component of this that I think is really useful because like anything else in life, it is one thing to learn something in the abstract and it is another thing to have to learn how to apply that.”


Wyze launches version 3 of its $20 security camera

TechCrunch, Frederic Lardinois


from

Wyze first made a name for itself when it launched its $20 indoor security camera a few years ago. Since then, the company branched out into other smart home products, ranging from doorbells to scales. Today, it’s going back to its origins with the launch of the Wyze Cam V3, the third generation of its flagship camera.

The new version is still $20 (though that’s without shipping unless there’s a free shipping promotion in the Wyze store), but the company redesigned both the outside and a lot of the hardware inside the camera, which is now also IP65 rated, so you can now use it outdoors, too.


Cryo–electron microscopy breaks the atomic resolution barrier at last

Science, Robert F. Service


from

If you want to map the tiniest parts of a protein, you only have a few options: You can coax millions of individual protein molecules to align into crystals and analyze them using x-ray crystallography. Or you can flash-freeze copies of the protein and bombard them with electrons, a lower resolution method called cryo–electron microscopy (cryo-EM). Now, for the first time, scientists have sharpened cryo-EM’s resolution to the atomic level, allowing them to pinpoint the positions of individual atoms in a variety of proteins at a resolution that rivals x-ray crystallography’s.

“This is just amazing,” says Melanie Ohi, a cryo-EM expert at the University of Michigan, Ann Arbor. “To see this level of detail, it’s just beautiful.” Because the heightened resolution reveals exactly how complex cellular machines carry out their jobs, improvements in cryo-EM should yield countless new insights into biology.


Experts link thousands of Minnesota COVID-19 cases to sports

FOX 9 (MN)


from

Minnesota health officials say they have connected more than 3,400 COVID-19 cases to sports, requiring 7,000 households to isolate.

In a news conference Monday, state infectious disease expert Kris Ehresmann said 593 of those cases have been traced to high school athletes and 309 more have been traced to middle school athletes.


Google collaborates with NOAA to use artificial intelligence for weather forecasting, research

TheHill, Rebecca Klar


from

Google and the National Oceanic and Atmospheric Administration (NOAA) have signed a three-year deal to use the tech giant’s artificial intelligence and machine learning to enhance the agency’s environmental monitoring, weather forecasting and climate research, according to a joint announcement released Tuesday.

Research under the deal initially focused on developing small-scale artificial intelligence and machine learning systems, and based on the results, NOAA and Google Cloud will focus on executing full-scale prototypes the agency could use across its organization.


DePaul Researchers Help Narrow Data Gap in COVID-19 Testing

WTTW, Chicago News, Alexandra Siles


from

When the pandemic hit, the city of Chicago found there was a significant information gap when trying to collect race and ethnicity data.

DePaul University has since developed a program that has narrowed the “unknown” race data gap in COVID-19 tests from 47% to 11%.

Joining “Chicago Tonight” to talk about how they did that and why collecting that kind of information is important are:

Daniela Stan Raicu, a professor of visual computing, artificial intelligence and bio-informatics at DePaul University’s School of Computing. She is also the director of DePaul’s Center for Data Science and is the associate provost for research. [video, 7:05]


The Science That Spans #MeToo, Memes, and Covid-19

WIRED, Ideas, C. Brandon Ogbunu


from

The technical term is “directed onion decomposition.”

It describes how centrally embedded an individual is in a network of others. The deeper in this “onion” they are, the more connections they have. The network being studied: NHL Hockey fights.

Researchers at the University of Vermont, the University of Colorado Boulder, and Dartmouth College analyzed 10 years of hockey fight data and reconstructed these brawls into a network where lines were drawn between participants. They found that hockey enforcers who were more centrally connected to others through combat tended to be stronger fighters.

Because “enforcers”—whose primary role is to protect their teammates, intimidate opponents, and fight—are a small proportion of hockey players, they provide a model for how network structure can reveal features of how people who participate in non-normative behaviors function in a “society.”


Even when algorithms outperform humans, people often reject them

University of Chicago, Booth School of Business, Chicago Booth Review, Jeff Cockrell


from

Data science has created more and better applications for algorithms, particularly those that use machine learning, to help predict outcomes of interest to humans. But has the progress of algorithmic decision aids outpaced people’s willingness to trust them? Whether humans will put their faith in self-driving cars, ML-powered employment screening, and countless other technologies depends on not only the performance of algorithms, but also how would-be users perceive that performance.

In 2015, Chicago Booth’s Berkeley J. Dietvorst, with University of Pennsylvania’s Joseph P. Simmons and Cade Massey, coined the phrase “algorithm aversion” to describe people’s tendency to distrust the predictions of algorithms, even after seeing them outperform humans. Now, further research from Dietvorst and Booth PhD student Soaham Bharti suggests that people may not be averse to algorithms per se but rather are willing to take risks in pursuit of exceptional accuracy: they prefer the relatively high variance in how well human forecasters perform, especially in uncertain contexts. If there’s a higher likelihood of getting very good forecasts, they’ll put up with a higher likelihood of very bad ones.


Machine learning helps hunt for COVID-19 therapies

Michigan State University, MSU Today


from

Michigan State University Foundation Professor Guowei Wei wasn’t preparing machine learning techniques for a global health crisis. Still, when one broke out, he and his team were ready to help.

The group already has one machine learning model at work in the pandemic, predicting consequences of mutations to SARS-CoV-2. Now, Wei’s team has deployed another to help drug developers on their most promising leads for attacking one of the virus’ most compelling targets. The researchers shared their intel Oct. 21 in the peer-reviewed journal Chemical Science.

Prior to the pandemic, Wei and his team were already developing machine learning computer models — specifically, models that use what’s known as deep learning — to help save drug developers time and money. The researchers “train” their deep learning models with datasets filled with information about proteins that drug developers want to target with therapeutics. The models can then make predictions about unknown quantities of interest to help guide drug design and testing.


Superheroes of Deep Learning Vol 1: Machine Learning Yearning

Approximately Correct blog, Falaah Arif Khan


from


Tools & Resources



How Kaggle Makes GPUs Accessible to 5 Million Data Scientists

NVIDIA Developer News Center, Megan Risdale


from

Engineers and designers at Kaggle work hard behind the scenes to make it easy for our 5 million data scientist users to focus on learning and improving their deep learning models instead of ML ops like environment setup and resource provisioning.

Kaggle’s community comes to the platform to learn and apply their skills in machine learning competitions. To do this, our users use Kaggle Notebooks, a hosted Jupyter-based IDE.


Careers


Full-time positions outside academia

Director or Deputy Director of Partnerships



The Carpentries; Remote

MSR NYC Fairlearn data science contractor position



Microsoft Research; New York, NY

Machine Learning Engineer



Development Seed; Washington, DC, or Lisbon, Portugal
Tenured and tenure track faculty positions

Tenure-Track Position in Machine Learning and Artificial Intelligence



Concordia University, Gina Cody School of Engineering and Computer Science; Montreal, QC, Canada

Computational Linguistics: Assistant Professor



Boston University, Department of Linguistics; Boston, MA

Biostatistics Faculty



George Mason University, Department of Global & Community Health; Fairfax, VA

Leave a Comment

Your email address will not be published.