NYU Data Science newsletter – September 12, 2016

GROUP CURATION: N/A

IN A narrow sense, the global financial crisis of 2008 was unprecedented. It was the result of a range of problems that had built up over time: light regulation of banks, overly complex credit products, tighter cross-border linkages and irrational exuberance in the housing market. But while that precise combination of factors had never been seen before, the trajectory from excessive risk-taking to financial chaos was a familiar one, whether to students of America’s volatile banking industry in the 19th century or to investors who remembered Asia’s woes in the late 1990s. Each crisis is unique, but meltdowns occur regularly enough that they exhibit certain patterns. What causes financial crises?

A mapping program, called PiCnIc for short, aims to help physicians in staying a step ahead of cancer and preparing long-term treatment plans with fewer elements of surprise.

How does it work?

PiCnIc “takes in patient data, guesses potential scenarios and tells you the most likely scenario,” says Bud Mishra, a professor at New York University’s Courant Institute of Mathematical Sciences who led a multinational team in developing the program.

Have you ever wondered what your credit score might taste like?

Numbers reflect interest rates, determine the outcome of presidential elections and dominate global commerce, and in turn they govern our lives. But algorithms are invisible and intangible. How can you make them appear more physical?

An innovative design project dubbed Data Cuisine has turned fuzzy maths into something you can bite into – literally. The process encourages people to examine topical issues by converting data-driven dishes into conversation pieces.

New York University biologist Elodie Ghedin will study the host response to Zika virus infections under a $1 million grant from the National Institute of Allergy and Infectious Diseases.

“The immune mechanisms associated with severe and mild disease induced by Zika virus are currently unknown—no biomarkers have yet been identified that are clearly associated with disease severity,” explains Ghedin, a professor in NYU’s Department of Biology and College of Global Public Health. “The goal of our research is to find key drivers of disease severity, but also to identify predictive biomarkers that will help identify patients that are at risk for developing neurological problems due to Zika.”

ING Bank’s main data center in Bucharest, Romania, was severely damaged over the weekend during a fire extinguishing test. In what is a very rare but known phenomenon, it was the loud sound of inert gas being released that destroyed dozens of hard drives. The site is currently offline and the bank relies solely on its backup data center, located within a couple of miles’ proximity.

In the last couple of years, deep learning techniques have transformed the world of artificial intelligence. One by one, the abilities and techniques that humans once imagined were uniquely our own have begun to fall to the onslaught of ever more powerful machines. Deep neural networks are now better than humans at tasks such as face recognition and object recognition. They’ve mastered the ancient game of Go and thrashed the best human players.

But there is a problem. There is no mathematical reason why networks arranged in layers should be so good at these challenges. Mathematicians are flummoxed. Despite the huge success of deep neural networks, nobody is quite sure how they achieve their success.

Today that changes thanks to the work of Henry Lin at Harvard University and Max Tegmark at MIT. These guys say the reason why mathematicians have been so embarrassed is that the answer depends on the nature of the universe. In other words, the answer lies in the regime of physics rather than mathematics.

The real value of digital curation comes from leveraging software to enable humans to quickly analyze and process a subset of the categorized data to determine the tone, narrative, and impact of the campaign or topic as a whole. The software offers access to the data, while humans extract unique, contextual elements of the data to make it useful and actionable. Through digital curation, the reporting of insights becomes more than just raw performance numbers on a campaign or topic. Results can be presented in a more persuasive way by presenting stakeholders with what consumers, media, and competitors are actually saying within the context of a topic – similar to a comment card.

The release next week includes 19 papers by Gaia astronomers who have already seen the data.
1. Where is the Milky Way’s dark matter?
2. Disputed stellar distances
3. Thousands of new worlds
4. How fast is the universe expanding?
5. Invisible asteroid threats
6. Milky Way archeaology

More space news:

Gaia space telescope plots a billion stars (September 14, BBC News, Jonathan Amos)

Andrew Connolly: What Data Will Be Discovered By The World’s Most Powerful Telescope? (9 September 2016, NPR: TED Radio Hour)

Hogg’s Research: #GaiaDR1 zero-day (September 14, David Hogg, Hogg’s Research blog)

The Pentagon’s research agency has a new challenge for scientists: make wireless radios with artificial intelligence that can figure out the most effective, efficient way to use the radio frequency spectrum, and win a pile of cash.

Winners of the Defense Advanced Research Projects Agency’s (DARPA) Spectrum Collaboration Challenge (SC2) could take home up to $3.5 million, but to do that, teams will have to demonstrate new technologies that represent a “paradigm shift” with both military and commercial applications, said Paul Tilghman, a DARPA program manager who is leading the challenge.

“With the proliferation of data sets from the explosion of the internet of things, demand for data scientists will grow exponentially,” said John Philosophos, a business development partner at Great Oaks Venture Capital LLC who lobbied Marquette to act on that demand. “This is an acute national problem, and it’s exciting to see Marquette on the forefront of the solution.”

Marquette began this fall offering an undergraduate data sciences major. Scheel, who is now a junior, happily signed on.

Big Data is everywhere — even the skies. Astronomer Andrew Connolly shows how large amounts of data are being collected about our universe, and how it will help lead to new discoveries. [audio, 13:09]

Also in astro last week, Gaia:

Milky Way mapper: 6 ways the Gaia spacecraft will change astronomy (September 09, Nature News & Comment)

Gaia space telescope plots a billion stars (September 14, BBC News)

Gaia’s Billion-star Map Hints at Treasures to Come (September 14, ESA)

Hogg’s Research: #GaiaDR1 zero-day (September 14, David Hogg, Hogg’s Research blog)

Boston, MA To apply to present at the battle, fill out the application form before September 26, 2016.

Barcelona, Spain This workshop will bring together researchers in machine learning, computer vision, robotics, computational neuroscience, and cognitive development to discuss artificial systems that capture or model intuitive physics by learning from footage of, or interactions with a real or simulated environment. — 9 December 2016

Deadline for submissions is Friday, 14 October 2016

Barcelona, Spain This workshop seeks to brings together experts in the fields of IML, reinforcement learning (RL), human-computer interaction (HCI), robotics, cognitive psychology and the social sciences to share recent advances and explore the future of IML. Some questions of particular interest for this workshop include: How can recent advancements in machine learning allow interactive learning to be deployed in current real world applications? How do we address the challenging problem of seamless communication between autonomous agents and humans?

Deadline for submissions is Friday, October 14.

Barcelona, Spain At this symposium, we will review the latest developments in all of these fields, and focus not only on RNNs, but also on learning machines in which RNNs interact with external memory such as neural Turing machines, memory networks, and related memory architectures. — 8 December 2016

Deadline for submissions is 15 October 2016.

Montreal, Canada The International AAAI Conference on Web and Social Media (ICWSM) is a forum for researchers from multiple disciplines to come together to share knowledge, discuss ideas, exchange information, and learn about cutting-edge research in diverse fields with the common theme of online social media. — 17-20 May 2016

Deadline for abstracts submissions is Friday, 16 January 2016.

“This post is Part 4 of the Deep Learning in a Nutshell series: Reinforcement learning. I use analogies and images whenever possible to provide easily digestible bits that make up an intuitive overview of the field of deep learning.”

At the Data Science Association our members often complain about the major data engineering problem of finding the right tools and programming models to build both robust data processing pipelines and efficient ETL processes for data transformation and integration.

Beam (incubating) attempts to solve this problem by providing a unified programming model to create data processing pipelines. The Apache Beam open source project is currently in incubation mode and we invite you to join the community and pitch in to help build.

You start by building a program that defines the pipeline using one of the open source Beam SDKs. The pipeline is then executed by one of Beam’s supported distributed processing back-ends, which include Apache Flink, Apache Spark, and Google Cloud Dataflow.

The PCC toolkit transforms consent from a signature on a legal form to a process that educates. Credit to Sage Bionetworks.

OpenPiton® is the open source release of the Princeton Piton many-core processor, which has been designed and taped-out in March 2015 in Princeton Parallel Group under supervision of Prof. David Wentzlaff.

Bailiwick allows reporters to quickly and efficiently uncover new investigative story ideas in campaign finance data. The system contains data for 2016 federal elections. Search for a race or a candidate that matters to you, or start with one of the races below. Log in to follow a candidate and receive alerts about story ideas and new filings related to a campaign.