Data Science newsletter – October 11, 2016

The very first Cybathlon in the world took place today. The event was completely sold out: some 4,600 visitors packed into the SWISS Arena Kloten to support the 66 teams from various countries. The pilots pitted their skills in six disciplines and demonstrated most impressively how novel technologies can assist people with disabilities in their daily life.

Ericsson has been a key driver of 5G deployment across Europe, working closely with the likes of the European Commission to identify and approach its opportunities and challenges. Its expertise will no doubt be invaluable when it comes to developing a unified European approach.

“We still need to understand much much more, as this will be different to mobile broadband,” adds Jonas Näslund, Head of Strategy at Ericsson’s Business Radio division. “We need to understand exactly what the industries need, and so far they have been very eager to explore the benefits and work with us.”

The CIA has found a way to predict social unrest, extremist activities, and other threats to national security almost a week in advance: Combine lots of data, machine learning, and other technologies with old school know-how. It’s not the future as imagined by Minority Report — no psychic abilities are involved — but it could eventually allow the agency to peer further and further into the future.

Speaking at the Next Tech event on Tuesday, CIA deputy director for digital innovation Andrew Hallman said that the agency is using these techniques to great effect in the Directorate for Digital Innovation, which was established in 2015 to help the CIA embrace new tech.

“We have, in some instances, been able to improve our forecast to the point of being able to anticipate the development of social unrest and societal instability some I think as near as three to five days out,” Hallman said at the event.

OpenTrials is a collaboration between Open Knowledge and Dr Ben Goldacre from the University of Oxford DataLab. It aims to locate, match, and share all publicly accessible data and documents, on all trials conducted, on all medicines and other treatments, globally. To find out more read this paper.

Explore the public beta version of OpenTrials here. [Video Walkthrough]

Comedian and broadcaster Iain Lee considers some of the matches made between news images and artworks, paired by #Recognition’s artificial intelligence. Can this comparison change the way we perceive art?

Safe when taken on their own, some prescription drugs become deadly when combined. Many of these interactions are well known, but others remain hidden to doctors, drug companies, and patients. Identifying these risky combinations has become a priority as the number of Americans on multiple medications rises.

Leveraging the power of big data, the researchers found a way to expand and improve the search for drug interactions. Mining a government database of reported drug side effects and a university hospital archive of patient records, they discovered eight pairs of drugs that are associated with a higher risk of a potentially deadly heart condition. Testing one of the pairs on individual heart cells in the lab, they discovered why the drug combination may disrupt the heart’s normal electrical activity in some patients.

Over a year before Swipes’ inception, NYU graduate student Jon Chin founded Share Meals after seeing a post made by a food-insecure student in the NYU Secrets Facebook group. A meal-sharing platform similar to Swipes, Share Meals became effective almost immediately after Chin launched the app.

“We operated only within the last seven days of that school year,” Chin said. “In that time we were able to match 400 students with swipes, and we actually had 400 additional swipes that were donated that we couldn’t find people to take. Even from the start, we had a huge surplus.”

At the 2016 IEEE 2nd International Conference on Collaboration and Internet Computing (CIC 2016) in Pittsburgh, Pennsylvania this November, Star Ying and I will present a paper on Big Data Privacy (read paper here).

In this paper, we provide a simple description of something that should be obvious to most?—?there is no privacy when it comes to Big Data.

Thanks to Tinder and OKCupid, it’s never been easier to find the love of your life on the Internet. But what happens after you find that person? It’s pretty much a crapshoot, because as of yet, there’s no magic formula to determine whether or not two people are actually romantically compatible.

The DNA matching startup Instant Chemistry, however, thinks it’s found that formula, and it’s as simple as expectorating into a tube. That’s why they’ve partnered up with the dating website Singld Out to offer DNA kits to users, to help determine how genetically compatible you are with a potential mate. All you have to do is fill out a brief personality assessment, send a saliva sample, and wait and see if you’re a match made in the genetic heavens.

Choosing the best treatment for a cancer patient is often an inexact science. Drugs that work well for some patients may not help others, and tumors that are initially susceptible to a drug can later become resistant.

In a new approach to devising more personalized treatments, researchers at MIT and Dana-Farber Cancer Institute have developed a novel way to test tumors for drug susceptibility. Using a device that measures the masses of single cells, they can predict whether a particular drug will kill tumor cells, based on how it affects their growth rates.

This election cycle has amplified the voices of several Young Turks whose desk-bound, scientific-method approach to journalism cuts against the grain of more traditional reporting techniques. Even more so than four years ago, 2016 has showcased a crop of political reporters who take a dramatically different approach from the boys-on-the-bus ethos of earlier generations stretching from David Broder to Hunter S. Thompson through Dan Balz. Sites like FiveThirtyEight, Politico, Slate, and The New York Times’ The Upshot have embraced statistical analysis as an antidote to anecdotal extrapolation.

The rise of this sort of data journalism has its critics. Silver “does not take a side, except the side of no side,” wrote Leon Wieseltier just after FiveThirtyEight launched at ESPN. “He does not recognize the calling of, or grasp the need for, public reason; or rather, he cannot conceive of public reason except as an exercise in statistical analysis and data visualization.” For those on Wieseltier’s side of the argument, there is a feeling that something essential is lost when human decisions are reduced to data points, that data journalism is in some way bloodless and empty.

For college students. Build models using baseball data and “think like a GM.”

Deadline to submit initial brief is Thursday, 20 ctober 2016.

On Wednesday, Center for Data Science’s Moore-Sloan fellow Michael Gill showcased his research on how war and conflict impact the economy.

The US Department of Defense outsources most of its military equipment. According to the data gathered about the DoD’s business contracts, they purchase tanks, planes, weapons, and other gear from almost 350,000 businesses and multinational corporations. Since these expenditures account for almost 1% of the global GDP annually, it is unsurprising to assume that military events would affect the economy. But is it possible to be more specific about this correlation?

Below I’ve listed every way I know of that people get paid for open source work, roughly ordered from small to large. Each funding category links to several real examples. (Wherever possible, I’ve tried to link to a useful article or page instead of just a homepage.)

Speaking of being open in a mostly different way, Jeffrey Leek released an eBook “How to be a modern scientist” [$10] on how to be ready for an alt-metric future in academia.

In Being a Scholar in the Digital Era, sociologists Jessie Daniels and Polly Thistlethwaite explain how to use social media as an opportunity for research, teaching, and engaging meaningfully in social justice.

Both make excellent cases that scholars and scholarship can thrive even if they venture out of the university onto Twitter, GitHub, FigShare, and the op-ed page of the local newspaper. Thank goodness.

My recommendations for scholarly social media this week?

Andrew Gelman’s stats blog, because it’s always on my list.

a 2013 blog post by Roger Peng on collaborating with scientists: “if you understand where the data come from (as in literally, the data come from this organ in this person’s body), then you might not be so flippant about asking for an extra 100 subjects.”

On cruelty and kindness in academia by Kelly J. Baker, “The increasing reliance on contingent labor and the decrease in tenure-track jobs makes kindness seem like a luxury” in academia.