NYU Data Science newsletter – July 21, 2016

GROUP CURATION: N/A

Ultimately, what the tech industry really cares about is ushering in the future, but it conflates technological progress with societal progress. And perhaps all of us have come to rely too deeply on machinery and software to be our allies without wondering about the cost, the way technology doesn’t fix problems without creating new ones.

More data + social movements:

Can Big Data Help Head Off Police Misconduct? (July 19, NPR, All Tech Considered)

Here’s what 29 million tweets can teach us about Brexit (July 20, The Washington Post, Monkey Cage blog; Alexandra Siegel and Joshua Tucker)

Tweeting Turkey, or how social media may have fundamentally changed the future of coups (July 19, The Washington Post, Monkey Cage blog; Joshua Tucker)

Another tragedy, another #PrayFor, but what does it really say about who cares for whom? (July 18, The Conversation, Drew Margolin)

To the dismay of government contractors, the United States Digital Service is gloriously hacking away in the VA and the Pentagon.

Analysis of more than 29 million tweets collected at NYU’s Social Media and Political Participation (SMaPP) Lab provides key insights into the success of the “leave” campaign, the surprising dominance of economic issues in the online debate, and the referendum’s increasingly global audience.

For almost five months on Twitter, ‘leave’ was more popular than ‘remain’

GSK launched has launched the study for rheumatoid arthritis patients in the US, who will participate using their iPhones.

The purpose of the Patient Rheumatoid Arthritis Data from the Real World (PARADE) study – is not to test a new medicine but to “bring patients into the center of research so we can better understand the disease and its impact” according to the company.

Flickr

In December 1964, over a single evening session in Englewood Cliffs, New Jersey, John Coltrane and his quartet recorded the entirety of A Love Supreme. This jazz album is considered Coltrane’s masterpiece – the culmination of his spiritual awakening – and sold a million copies. What it represents is all too human: a climb out of addiction, a devotional quest, a paean to God.

Five decades later and 50 miles downstate, over 12 hours this April and fuelled by Monster energy drinks in a spare bedroom in Princeton, New Jersey, Ji-Sung Kim wrote an algorithm to teach a computer to teach itself to play jazz. Kim, a 20-year-old Princeton sophomore, was in a rush – he had a quiz the next morning. The resulting neural network project, called deepjazz, trended on GitHub, generated a buzz of excitement and skepticism from the Hacker News commentariat, got 100,000 listens on SoundCloud, and was big in Japan.

Big Data has been considered an essential tool for tech companies and political campaigns. Now, someone who has handled data analytics at the highest levels in both of those worlds sees promise for it in policing, education and city services.

For example, data can show that a police officer who has been under stress after responding to cases of domestic abuse or suicide may be at higher risk of a negative interaction with the public, data scientist Rayid Ghani says. [audio, 7:39]

Can Silicon Valley Really Do Anything to Stop Police Violence? (July 19, The New York Times Magazine, Jenna Wortham)

()

Technology Is Monitoring the Urban Landscape (July 20, The New York Times)

Facing realities – Face-recognition technology has suddenly become much more powerful. (July 14, The Economist, 1843 magazine)

Big City is watching you.

It will do it with camera-equipped drones that inspect municipal power lines and robotic cars that know where people go. Sensor-laden streetlights will change brightness based on danger levels. Technologists and urban planners are working on a major transformation of urban landscapes over the next few decades.

Much of it involves the close monitoring of things and people, thanks to digital technology. To the extent that this makes people’s lives easier, the planners say, they will probably like it. But troubling and knotty questions of privacy and control remain.

The U.S. Treasury Department’s research arm, Office of Financial Research, on Wednesday launched an online tool that allows investors to analyze the securities held by money market mutual funds.

The introduction of OFR’s “Money Market Monitor” comes at time when regulators have tightened oversight of the $2.7 trillion industry following the collapse of the Reserve Primary Fund in September 2008 at the height of the global financial crisis when its share price fell below $1.

Baidu, China’s internet search giant, has shown just what you can learn when you have access to enough location data.

The firm’s Big Data Lab in Beijing has announced that it has used billions of location records from its 600 million users as a lens on the Chinese economy, tracking the flux of people around offices and shops as a proxy measurement for employment and consumption activity. The lab even used the data to predict Apple’s second quarter revenue in China.

Can intelligent machines really match human creativity? Surely when we think artist, we think human and when we think artificial intelligence, we think machine. At least in the art field, we cannot seem to come to terms with the possibility of a machine truly making art. Why?

Also in data + art:

There is no difference between computer art and human art (July 20, Aeon Ideas, Oliver Roeder)

SIGGRAPH 2016 Art Gallery (begins July 24, SIGGRAPH)

Rain Room, by Random International (began July 14, The Los Angeles County Museum of Art)

Research suggests that hashtag use is governed by “complex contagion” dynamics, in which people wait to see if multiple friends have adopted a hashtag before adopting it themselves. This convergence can emerge naturally from the desire to meet others’ expectations. The more people see sympathy expressed with #PrayFor, the more they expect #PrayFor is where others will look for sympathy. And so the more they direct their messages to it.

Now that #PrayFor has “self-organized” into a convention, people treat it more like a reflection of a single, collective voice, rather than as an amalgamation of individual behaviors.

The Data Science Institute’s newest working group —Frontiers in Computing Systems—will try to address some of the bottlenecks facing scientists working with massive data sets at Columbia and beyond. From astronomy and neuroscience, to civil engineering and genomics, major obstacles stand in the way of processing, analyzing and storing all this data.

“We don’t have two years to process the data,” said Ryan Abernathey, a physical oceanographer at Columbia’s Lamont-Doherty Earth Observatory. “We’d like to do it in two minutes.”

Throughout the three days, I had conversations with about 40 chairs and deans, ranging from those at prestigious institutions such as Harvard, Yale, Princeton, and Northwestern, to the long tail of excellent CS workhorses such as Texas A&M, Purdue, Harvey Mudd, CU Boulder, and UC Santa Cruz. I also connected with chairs more local to the University of Washington, including Joe Sventek at the University of Oregon and James Hook at Portland State University. This was a networking fantasy for anyone interested in the infrastructure of academia and research policy, not to mention an incredible chance to have substantive research and policy conversations across computing with some of the best researchers in the world.

I learned much about the life of computing leaders. For example, most chairs and deans, at the most basic level, are necessarily concerned with the slow building of academic infrastructure.

Designers are constantly asked to design for the future—a future that by its nature is a moving target. The O’Reilly Design Conference will provide designers with the skills, connections, and inspiration they need to shape the products and services of today and tomorrow.

San Francisco, CA Sunday-Wednesday, March 19-22.

Deadline for speaking proposals is Wednesday, September 7.

The experience and intuition of a human doctor are crucial to quality healthcare, but data science can be used to help doctors make better decisions.

Eric Williams, Vice President of Data Science and Analytics at Omada Health, says his team is using data science to facilitate behavioral change—specifically weight loss—in chronic disease prevention and treatment. Williams said that technological advancements have eased data collection and allowed Omada to accurately link weight loss to effective treatments.

Recent years have seen a widespread adoption of machine learning in industry and academia, impacting diverse areas from advertisement to personal medicine.

As more and more areas adopt machine learning and data science techniques, the question arises on how much expertise is needed to successfully apply machine learning, data science and statistics.

Following our announcements from GCP NEXT in March, we’re excited to share updates about Cloud Platform expansion and machine learning. Today we’re launching two new Machine Learning APIs into open beta and expanding our footprint in the United States.

We agreed to customize the California Digital Library’s (CDL) DMPTool (data management plan tool), to which Harvard subscribes, but at that point had only been used in its generic form. Harvard Library is a point of contact for Harvard-affiliated researchers University-wide seeking data management support and services. A Working Group (WG) under HL’s Stewardship Standing Committee was convened in Fall 2015 to roll-out a customized version of DMPTool by early 2016. … [The WG] decided the best way of supporting the tool and to better communicate the benefits of research data management was to link the customized tool to a topical LibGuide.

One of the difficult problems in using machine learning to generate sequences, such as melodies, is creating long-term structure. Long-term structure comes very naturally to people, but it’s very hard for machines. Basic machine learning systems can generate a short melody that stays in key, but they have trouble generating a longer melody that follows a chord progression, or follows a multi-bar song structure of verses and choruses. Likewise, they can produce a screenplay with grammatically correct sentences, but not one with a compelling plot line. Without long-term structure, the content produced by recurrent neural networks (RNNs) often seems wandering and random.

But what if these RNN models could recognize and reproduce longer-term structure? Could they produce content that feels more meaningful – more human? Today we’re open-sourcing two new Magenta models, Lookback RNN and Attention RNN, both of which aim to improve RNNs’ ability to learn longer-term structures. We hope you’ll join us in exploring how they might produce better songs and stories.