NYU Data Science newsletter – June 3, 2016

GROUP CURATION: N/A

“NSF has been a leader in supporting fundamental research and development in data analysis methods, techniques and tools that enhance the benefits brought to us by the data and information revolution,” said France Córdova, NSF director. “The Federal Big Data Research and Development Strategic Plan lays out opportunities for further collaboration among Federal agencies, academia, government and industry that will catalyze a Big Data innovation ecosystem and harness the benefits of Big Data for years to come. The strategy will help guide NSF’s future investments in this area of critical national importance.”

Nearly 250 million video surveillance cameras have been installed throughout the world, and chances are you’ve been seen by several of them today. Most people barely notice their presence anymore — on the streets, inside stores, and even within our homes. We accept the fact that we are constantly being recorded because we expect this to have virtually no impact on our lives. But this balance may soon be upended by advancements in facial recognition technology.

Also, in facial recognition & computer vision:

Big Data: Facial Recognition and the Biometrics Movement (May 31, MapR Converge blog)

Undercovered: Facial Recognition and the Future of Privacy (May 17, Mediaite)

Deep Learning Trends @ ICLR 2016 (June 01, Tomas Malisiewicz, Tombone’s Computer Vision Blog)

International Conference on Learning Representations (ICLR) 2016, San Juan (May 30, VideoLectures.NET)

The following is an excerpt from a blog post written by Daniel Turner about the International Congress of Qualitative Inquiry Conference, attended by NYU Data Services’s Sarah DeMott:

“I was lucky enough to attend the ICQI 2016 conference last week in Champagne at the University of Illinois. We managed to speak to a lot of people about using Quirkos, but there were hundreds of other talks, and here are some pointers from just a few of them!

Most disorders are caused by a combination of multiple genetic and/or environmental factors. If two diseases are caused by the same molecular mechanism, they tend to co-occur in patients. Here we provide a quantitative method to disentangle how much genetic or environmental risk factors contribute to the pathogenesis of 358 individual diseases, respectively. We pool data on genetic, pathway-based, and toxicogenomic disease-causing mechanisms with disease co-occurrence data obtained from almost two million patients. From this data we construct a multilayer network where nodes represent disorders that are connected by links that either represent phenotypic comorbidity of the patients or the involvement of a certain molecular mechanism. From the similarity of phenotypic and mechanism-based networks for each disorder we derive measure that allows us to quantify the relative importance of various molecular mechanisms for a given disease. We find that most diseases are dominated by genetic risk factors, while environmental influences prevail for disorders such as depressions, cancers, or dermatitis. Almost never we find that more than one type of mechanisms is involved in the pathogenesis of diseases.

We examine some economic impacts of hospital admissions using an event study approach in two datasets: survey data from the Health and Retirement Study, and hospital admissions data linked to consumer credit reports. We report estimates of the impact of hospital admissions on out-of-pocket medical spending, unpaid medical bills, bankruptcy, earnings, income (and its components), access to credit, and consumer borrowing. The results point to three primary conclusions: non-elderly adults with health insurance still face considerable exposure to uninsured earnings risk; a large share of the incremental risk exposure for uninsured non-elderly adults is borne by third parties who absorb their unpaid medical bills; the elderly face very little economic risk from adverse health shocks.

Even before looking at this paper I was positively disposed toward it for two reasons. First because I do think there are incentives that encourage scientists to follow the forking paths toward statistical significance and that encourage journalists to publish this sort of thing. And I also see incentives for scientists and journals (and even the Harvard University public relations office; see the P.P.S. here) to simply refuse to even consider the possibility that published results are spurious.

[1605.09511] The Natural Selection of Bad Science (May 31, arXiv, Physics > Physics and Society; Paul E. Smaldino, Richard McElreath)

When Adrian Ludwig describes the ideal approach to computer security, he pulls out an analogy. But it’s not a lock or a firewall or a moat around a castle. Computer security, he says, should work like the credit card business.

A credit card company, he explains, doesn’t eliminate risk. It manages risk, using data describing the market as a whole to build a different risk profile (and a different interest rate) for each individual. Computer security, Ludwig believes, should work in much the same way. “The model of good and bad—white and black—that the security community prescribes?” he says. “It’s going to be all black unless we accept that there are going to be shades of gray.”

This is pretty much what you’d expect him to say. Ludwig works at Google, where he oversees security for Android.

The SocInfo 2016 Committee invites proposals for Workshops Day at the Eighth International Conference on Social Informatics (SocInfo 2016).

Seattle, WA The Workshops Day will be held on Monday, 14 November.

Deadline for workshop submissions is Friday, July 1.

In his research, Panos Ipeirotis—an Associate Professor at the Center for Data Science and the Stern School of Business—combines the computing power of machines with the problem solving capabilities of humans, as a way of generating more effective machine learning techniques. He recently published a paper titled, “Beat the Machine: Challenging Humans to Find a Predictive Model’s ‘Unknown Unknowns,” which examined the effectiveness of having humans look for the blind spots in machine learning techniques. … We spoke with Panos about his research concerning the “unknown unknowns” of machine learning.

Today we are very happy to announce that the 1.0 alpha of both our products -PubSweet [for science blogging] and INK [for managing file conversions] are available.

Learn the ways of Linux-fu, for free.