Data Science newsletter – September 27, 2018

GROUP CURATION: N/A

… In recent work, [Christopher] Tonetti and Charles I. Jones, also an economics professor at Stanford GSB, studied how data is valued, with an eye toward determining what an ideal market ought to look like — a question that economists, until this point, have given limited consideration.

“Who should own data?” they ask in their published working paper. “What restrictions should apply to the use of data?”

In answering these questions, they found that the current arrangement is far from optimal: People, not companies, should have rights to their data, and people, not companies, should be able to sell it as they see fit.

Boston University is proposing a 19-story tower that would house a data sciences center in the heart of its Charles River campus.

The 350,000-square-foot tower would be built at 645-665 Commonwealth Ave., currently a surface parking lot at the corner of Commonwealth Avenue and Granby Street, a BU official wrote in a Sept. 25 letter of intent to the Boston Planning and Development Agency.

The University of Minnesota received a three-year, $1.43 million grant earlier this month from the National Science Foundation to further advance machine learning techniques to better monitor global agriculture and environmental changes.

Machine learning is when computers can “learn” from data, without needing direct human programming. Through this project, it can be used to help society address climate change issues, manage land use and natural resources and sustainably feed a growing population.

Darktrace differentiates itself in using advanced machine learning and mathematics developed at the University of Cambridge to identify abnormalities in a company’s IT network that might be an attack.

The company also aims to tackle the latest phenomenon of artificial intelligence-based cyber attacks with its own AI-based software.

A venture fund at Minnetonka-based UnitedHealth Group is part of a $33 million fundraising round for an Iowa startup that’s developing an autonomous diagnostic system featuring artificial intelligence (AI).

Called IDx, the company this year received clearance from the Food and Drug Administration to market the first medical device using AI to detect greater than a mild level of the eye disease diabetic retinopathy in adults with diabetes.

The startup on Wednesday announced fundraising that’s intended to accelerate market adoption of the technology and development of related products.

Silicon Valley-area big data analytics company Palantir Technologies, founded by Peter Thiel, has signed a $7 million contract with the National Institutes of Health to supply its software platform to the organization.

Specifically, the National Center for Advancing Translational Sciences (NCAT) and its related groups will be using the tech to automatically aggregate research data from public and private sources into a single interface, allowing for more streamlined analysis.

The Open Data Initiative, launched today by Microsoft, Adobe and SAP, aims to help firms to extract more value from their data to power AI-driven projects, improve privacy and provide a broader spectrum of real-time insights.

The three software giants unveiled the new project in an announcement at the Microsoft Ignite conference in Orlando, FL.

The Open Data Initiative (ODI) will involve the three partners boosting the data exchange and interoperability between their various platforms and applications through a common data model. These platforms will include: Microsoft Dynamics 365, Adobe Experience Platform, Adobe Experience Cloud, S/4HANA and SAP C/4HANA.

One of the many ways that ads get in front of your eyeballs on Facebook and Instagram is that the social networking giant lets an advertiser upload a list of phone numbers or email addresses it has on file; it will then put an ad in front of accounts associated with that contact information. A clothing retailer can put an ad for a dress in the Instagram feeds of women who have purchased from them before, a politician can place Facebook ads in front of anyone on his mailing list, or a casino can offer deals to the email addresses of people suspected of having a gambling addiction. Facebook calls this a “custom audience.”

In a companywide meeting earlier this year, Mark Zuckerberg was asked if Instagram could have hit one billion users if it had not been bought by Facebook.

Probably not, he said. At least, not as quickly.

But at a later meeting a mile or so down the road at Instagram’s headquarters in Menlo Park, Calif., a few streets, interestingly enough, off a thoroughfare called Independence Drive, the popular photo-sharing app’s co-founders, Kevin Systrom and Mike Krieger, had a slightly different answer to that question.

Perhaps. Eventually.

Federal agencies would have to get serious about adopting artificial intelligence under a bill introduced Wednesday.

The Artificial Intelligence in Government Act, sponsored by Sens. Brian Schatz, D-Hawaii, Cory Gardner, R-Colo., Kamala Harris, D-Calif., and Rob Portman, R-Ohio, would provide resources for feds to start exploring how AI could play into their agencies.

The bill would require the General Services Administration to bring more AI experts on board and conduct original research on federal AI policy. It would also establish a board of experts who would advise agencies on AI implementation and help them overcome obstacles to adoption.

A toxic red tide, or harmful algae bloom, is killing swaths of marine life and affecting the health of people living along Florida’s southwest coast. Nationwide, harmful algae blooms cost an estimated $50 million each year. Excess nutrients such as nitrogen and phosphorus flowing downstream act as fertilizer, sparking these blooms in waterbodies such as the Gulf of Mexico, Lake Erie and Chesapeake Bay.

Paul Leisnham of the University of Maryland, College Park, is working to find out whether better management of nutrient-laden stormwater in urban areas can help. Leisnham has received one of 13 new grants from the National Science Foundation’s (NSF) Dynamics of Coupled Natural and Human Systems (CNH) program. Total funding for the 2018 CNH awards is $15 million.

Life insurance company John Hancock made a splash last week with the news that soon all its policies would come bundled with the option to let the company track your fitness — via either a website and app, or through the use of a fitness tracker like an Apple Watch or Fitbit.

The move underscores how fitness tracker data is an as-yet largely untapped gold mine for businesses — particularly in industries like insurance, whose financial bottom line directly depends on the health of their customers. John Hancock isn’t particularly shy about this: “The longer people live, the more money we make,” as the company’s CEO, Brooks Tingle, put it to the New York Times.

The published research on Fitbits and similar devices, however, has yet to uncover a clear link between fitness tracking and fitness, to say nothing of longevity and mortality, or of insurance companies’ profits. But there is some solid evidence that if the use of the devices is paired with incentives like rewards, challenges and leaderboards (“gamified,” in social science parlance) people can see real health benefits. It’s probably no accident, then, that the John Hancock policies lean heavily on those kinds of incentives.

Digital transformation is remaking the human world, but few are satisfied with how that’s been going. That’s especially true in media, where the dominant model of targeted advertising derived from data surveillance and used to fund free-to-the-public services like social media and search is increasingly viewed as unsustainable and undesirable.

Yale University is launching a new Master of Science in Health Informatics. The program will be housed in the Department of Biostatistics at the School of Public Health and directed by Professor Cynthia Brandt, M.D., Ph.D.

Applications are now being accepted for students interested in matriculation in the Fall of 2019.

A proposed CMS rule on health IT interoperability has been submitted for OMB review, following ONC’s long-awaited rule defining acceptable information blocking, which was sent over to the White House earlier last week. Once these two rules emanating from the December 2016 21st Century Cures Act emerge — it could be a matter of days, weeks or at most a couple months — we’ll have the big picture of the Trump administration’s plans for regulating health IT.

… To date, that image has been a bit blurry, although the general outlines are there. ONC Chief Don Rucker has made clear he is not a big fan of government regulation and at times has given the impression he’d rather not have to define information blocking at all. Over at CMS, Administrator Seema Verma has seemed somewhat more aggressive. Her agency has focused on trying to ensure greater patient access to data under its MyHealthEData initiative, but at the same time it has indicated a willingness to cut health care networks out of the Medicare program if they don’t accept interoperability mandates.

Berkeley, CA October 10, starting at 4 p.m., IRLE Director’s Room (2521 Channing Way). Speaker: John Voorheis, US Census Bureau.

New York, NY October 3, starting at 6 p.m., Citi (388 Greenwich Street, 27th Floor). “In this panel, Vera will convene some of the leading commentators on technology and justice reform to tackle these important issues.” [rsvp requested]

The author claims that the breakeven cost is ~ 2 months for single GPU vs AWS, and 2 weeks for the 4 GPU version, though one has to be careful with choosing components that will support well the 4 GPU version. [103 comments]

Some of you have probably heard about HDR, or High Dynamic Range, which can produce a higher level of brightness, a wider range of colors, and an increased level of detail and accuracy that empowers visual storytellers to deliver more impactful experiences for people to enjoy. Netflix has been delivering HDR video for several years now, so it made sense for us to start exploring ways to provide the benefits of HDR to the imagery we showcase in the browse experience where viewers initially discover and connect with stories.

We’re excited to roll out experimental HDR images for the very first time to the Netflix app on the latest generation of game consoles.

“We’ve created a general package for scikit-learn decision tree visualization and model interpretation, which we’ll be using heavily in an upcoming machine learning book (written with Jeremy Howard).”

The post goes into excruciating detail as to what you need in order to roll out a fairly decent cybersecurity program that attempts to meet all SEC cybersecurity requirements. I do not recommend sitting down and reading this in one sitting.

After being accepted for publication, each of Scientific Data’s Data Descriptors undergoes a curation process. During this process our in-house curators work with the authors to check that the associated data records are accurately cited, and that the provenance of the resulting data files is clear.

The most tangible output of the curation process is the machine-accessible metadata record that forms a part of each Data Descriptor (in ISA-Tab format1; http://isa-tools.org).