Data Science newsletter – October 4, 2016

The Center for Open Science and NYU’s Data Services are excited to announce the launch of OSF for Institutions at NYU. The Center for Open Science is a non-profit technology company that provides free and open-source services to increase inclusivity and transparency of research.

When Jonas Ranstam wakes up in the morning, one of the first things he does is review other people’s scientific papers. By his count, the independent medical statistician spends 24 hours a week peer reviewing—and in sum, over the past year, he has reviewed 661 papers. By one count, this is the most of any other scientific reviewer, and the dedication has won him the grand prize in the inaugural Sentinel of Science Awards. The prize, which involved a small amount of cash and is sponsored by a handful publishers, is being presented Friday by Publons, a platform that tracks reviewers’ progress. (Ranstam uses it himself.)

In addition to Ranstam, the Sentinel of Science Awards is also honoring top reviewers by country and field. The company hopes the awards will provide some recognition for what is often a thankless task. (It’s also a promotional tactic—there’s an award for “evangelist editors” who have referred the most people to the platform.) But recognizing peer review is a worthy cause: The labor it takes to peer review science papers would cost a couple billion dollars a year, if the reviewers were paid for their time, which they almost never are.

Amazon has hired a top eBay artificial intelligence executive to be its AI director at its A9 labs in Palo Alto.

Hassan Sawaf had held a number of positions at eBay (NASDAQ:EBAY) over the past three years, most recently listing his job there on his LinkedIn page as “head of artificial intelligence.” He also set up its cognitive computing group, helping shoppers searching in foreign languages find items listed in English.

NVIDIA is the inventor of the GPU, the computer device that creates interactive graphics on laptops, workstations, mobile devices, notebooks, and personal computers. It is Tesla’s supplier of choice for the components that power its 17? touchscreen in the Model S. NVIDIA is a global company with interests in many emerging computer technologies, one of which is autonomous driving systems.

A team of NVIDIA engineers working out of a former Bell Labs office in New Jersey decided to use deep learning to teach an automobile how to drive. As it happens, Bell Labs is where the earliest research into deep learning — sometimes called artificial intelligence or AI — took place.

I have been to a lot of hackathons, and I’ve organized a fair few too. Even though I was only at this event for a few hours on Sunday I can say hands down that it was the best organized hackathon that I’ve ever been to.

The key factor towards success was the groundwork that the DataKind UK “data ambassadors” did in the weeks leading up to the hackathon. They worked closely with the non-profits to prepare datasets that were ready to be worked on from the start, and they worked to help define questions that were important to the non-profits. I saw a lot of discipline from the teams and an amazing ability not to get sidetracked! I thought the results were phenomenal, and the reaction from the non-profits representatives was extremely positive.

Seattle and Vancouver are like fraternal twins separated at birth. Both are bustling Pacific Northwest coastal cities with eco-conscious populations that have accepted the bargain of dispiriting weather for much of the year in exchange for nearby ski slopes and kayaking and glorious summers.

Yet 140 miles of traffic-choked roads and an international border divide the two cities, keeping them farther apart than their geographic and cultural identities would suggest.

Now the political, academic and tech elite of both cities are looking for ways to bring them closer together, with the aim of continuing the growth of two of the most vibrant economies in North America.

Daily weather forecasts in the U.S. wouldn’t be nearly as accurate as they are without the three geostationary weather satellites that are parked 22,000 miles above Earth. Next month these predictions will get even better: the U.S. National Oceanic and Atmospheric Administration and NASA plan to launch the first of four satellites that should deliver what the agencies call “game-changing” capabilities for predicting both ordinary weather and dangerous storms such as hurricanes. These next-generation spacecraft are needed to replace the existing weather satellites, one of which reaches the end of its operational lifetime this year.

The new Geostationary Operational Environmental Satellite-R (GOES-R) will have the ability to scan for signs of rain, snow and lightning in the clouds above the entire continental U.S. every five minutes and to take smaller, focused images of problem areas every 30 seconds.

Researchers at the National Institute of Standards and Technology (NIST) may have changed the technology paradigm for DNA sequencers in their proposal for an entirely new material architecture that would represent the first DNA sequencer based on sensing motion in the membrane as the DNA thread passes through it.

In research described in the journal ACS Nano, the NIST researchers proposed a device in which a nanoscale ribbon of molybdenum disulfide is suspended over a metal electrode immersed in water. In this arrangement, the molybdenum disulfide acts as a kind of capacitor, storing an electrical charge. When a single strand of DNA is passed through a pore in the membrane, the membrane only flexes when a DNA base pairs up with and then separates from a complementary base affixed to the hole. It is this flexing that the motion sensor detects as an electrical signal.

In August, Illinois Circuit Court Judge James Eder made a quiet, yet important, ruling.

The small town of Altamont, Illinois (population: 2,304), had previously imposed a $1,500 fee on telecommunications business Metro Communications for laying Internet cables under the city’s public right-of-way. The justification was that the roll-out necessitated closing streets and digging, forcing the city to pay engineers as well as various administration fees.

As City Clerk Sarah Stephen put it, “The taxpayers should not have to absorb engineering fees to protect them from a company that’s going to come here and charge subscriber fees where they’re gonna make money.” Eder didn’t agree, saying the city broke the law.

Experts from across the data world came together in New York for Strata + Hadoop World in New York 2016. Below you’ll find links to highlights from the event.

1. The new dynamics of big data

In this exclusive article, Freedman shares the three biggest (often company-ending) mistakes startups make when it comes to setting up their systems:

They land themselves in Cloud Jail.

They get sucked in by “hipster tools.”

They don’t design for monitorability.

Earlier this month, on the 15th anniversary of the 9/11 attacks, the lower tip of Manhattan was thronged with soldiers in uniform, firefighters marching with photos of lost friends pinned to their backpacks, and tourists bumbling around the new mall at the World Trade Center. Firetrucks and police cars ringed Zuccotti Park and white ribbons adorned the iron fence around the churchyard on Broadway. Trash cans were closed up, with signs announcing “temporary security lockdown.”

So it felt a bit risky to be climbing up a street pole on Wall Street to closely inspect a microwave radar sensor, or to be lingering under a police camera, pointing and gesturing at the wires and antenna connected to it. Yet it was also entirely appropriate to be doing just that, especially in the company of Ingrid Burrington, author of the new book Networks of New York: An Illustrated Field Guide to Urban Internet Infrastructure, which points out that many of the city’s communications and surveillance programs were conceived and funded in response to the attacks.

Barcelona, Spain NIPS 2016 Workshop. Deadline for submissions to Learning Methods for Dialogue is Sunday, October 9.

More NIPS 2016 Workshops (submission deadline, workshop date):

Recurrent Neural Networks Symposium (15 October 2016, 8 December 2016)

Future of Interactive Learning Machines Workshop (14 October, 9 December)

NIPS Workshop on learning intuitive physics (14 October, 9 December)

Workshop on Machine Learning for Health (28 October, 9 December)

Approximate Inference (1 November, 9-10 December)

Interpretable ML for Complex Systems (20 October, 9 December)

Neural Abstract Machines & Program Induction (14 October, 10 December)

End-to-end Learning for Speech and Audio Processing Workshop (20 October, 10 December)

Machine Learning and the Law (3 November, 8 December)

“The Deep Learning AMI is a supported and maintained Amazon Linux image provided by Amazon Web Services for use on Amazon Elastic Compute Cloud (Amazon EC2).” … “It includes popular deep learning frameworks, including MXNet, Caffe, Tensorflow, Theano, and Torch, as well as packages that enable easy integration with AWS.”