NYU Data Science newsletter – May 23, 2016

GROUP CURATION: N/A

… Though hotly debated in privacy sectors, experts agree that dozens of tiny cameras and eventually nano cameras will soon be built into objects, providing devices the ability to see from every angle and in real time.

As Serge Belongie, Professor of Computer Vision at Cornell Tech points out, “There may be significant barriers to social acceptance faced by always-on cameras everywhere. People may ask, what’s the point? Where’s the value? Why would I want cameras pointed at me all of the time?

“While I don’t know when the turning point of acceptance will occur, it will coincide with a broad realization that this kind of visual technology, combined with smart cameras in fixed locations at home and at work, can in fact improve our health, aid our memory, and provide us new ways to care for loved ones.”

One of the major impediments in human aging research is the absence of a comprehensive and actionable set of biomarkers that may be targeted and measured to track the effectiveness of therapeutic interventions. In this study, we designed a modular ensemble of 21 deep neural networks (DNNs) of varying depth, structure and optimization to predict human chronological age using a basic blood test. To train the DNNs, we used over 60,000 samples from common blood biochemistry and cell count tests from routine health exams performed by a single laboratory and linked to chronological age and sex.

Google announced something for everyone yesterday at its 10th annual I/O developer conference. There were more details of a new version of Android; new messaging and video-calling apps; a built-in new VR platform for Android; and a good-looking Amazon Echo-like smart speaker called Google Home.

There was even a cool new research project called Instant Apps that will let users run portions of apps from the web without installing them first.

But the biggest theme stressed by Google CEO Sundar Pichai and his lieutenants, over and over again throughout the two-hour keynote, was that Google is doubling down on artificial intelligence as the next great phase of computing. And they believe Google can do it better than anyone else.

Martha Kim (Columbia University) & Hannah Wallach (Microsoft Research)

The Institute for the Interdisciplinary Study of Decision Making is excited to welcome Professor Catherine Hartley to its faculty. Professor Hartley is no stranger to the NYU community, having obtained her Ph.D. in psychology from here. She will begin this July as the Assistant Professor of Psychology.

Artificial intelligence (AI) via predictive analytics is a game-changer. At last week’s Kentucky Derby, an AI platform that had previously predicted the winners of the Oscars and Super Bowl predicted the Kentucky Derby superfecta. The AI platform predicted the first-, second-, third- and fourth-place horses at 540-1 odds, netting the technology’s inventor Louis Rosenberg $10,842 on a $20 bet. How will this translate to the federal government?

On May 3, The White House issued a very interesting document: “Preparing for the Future of Artificial Intelligence.” Recognizing that AI is a technology area full of both promises and potential perils, the White House Office of Science and Technology Policy announced that it will co-host four public workshops topics in AI to spur public dialogue on AI and machine learning.

Harvard’s Gary King spoke at The Post’s 2016 Transformers live journalism event May 18 in Washington, D.C. … “Big data is not actually about the data. The revolution is not that there’s more data available. The revolution is that we know what to do with it now. That’s really the amazing thing.

As specialists gather in private to discuss a grand plan for constructing a human genome, Drew Endy and Laurie Zoloth argue that such an enormous moral gesture should not be discussed behind closed doors.

Netflix is the world’s leading internet television network. That didn’t happen by accident or simple fortune – we are data-driven as part of our culture, and have built the tools needed to navigate the unchartered waters of delivering internet video at scale and becoming the first truly global storyteller in movies and television. [video, 18:48]

Advanced computing capabilities are used to tackle a rapidly growing range of challenging science and engineering problems, many of which are compute- and data-intensive as well. Demand for advanced computing has been growing for all types and capabilities of systems, from large numbers of single commodity nodes to jobs requiring thousands of cores; for systems with fast interconnects; for systems with excellent data handling and management; and for an increasingly diverse set of applications that includes data analytics as well as modeling and simulation. Since the advent of its supercomputing centers, the National Science Foundation (NSF) has provided its researchers with state-of-the-art computing systems. The growth of new models of computing, including cloud computing and publically available by privately held data repositories, opens up new possibilities for NSF. [118 page report, pdf]

President Obama wants to hear from YOU – kid scientists and innovators across the country – about what we can do to help shape the future of science, discovery, and exploration.

Also, from the White House:

Administration Issues Strategic Plan for Big Data Research and Development (May 23, The White House, Keith Marzullo)

Stanford Medicine X and Symplur are pleased to announce a joint initiative designed to spark scholarly research activity in healthcare social media. The Stanford Medicine X | Symplur Everyone Included™ Research Challenge seeks to encourage all health care stakeholders to collaborate on health care social media research.

Deadline for submissions is Friday, June 17.

Thanks to generous sponsorship from the R Consortium, Software Carpentry is running a two-day R instructor training class in Cambridge, UK, on September 19-20, 2016. If you are active in the R and/or Software and Data Carpentry communities, and wish to take part in this training, please fill in this application form.

We will select applicants, and notify everyone who applied, by June 30, 2016; those who are selected will be responsible for their own travel and accommodation.

Each year, millions of people suffer brain-related disorders and injuries and as a result, many face a lifetime of impairment with limited treatment options. This competition is based on one of the greatest challenges in neuroscience today – how to interpret brain signals. … Through this competition, you will play a key role in bringing the next generation of care to patients through advancing neuroscience research. Build the most intelligent model and accurately predict the image shown to a person based on electric signals in the brain.

Deadline for submissions is Friday, July 1.

We released a new version (0.5) which includes an interactive visualization, support for mobile phone recharges, support for Python 3, and clustering algorithms to handle both antenna and GPS locations. The computations are significantly faster and the memory footprint is reduced. This release is available on GitHub and PyPI.

Reddit is a social news website that has become a major driver of traffic to blog posts, videos, images and news articles. Alastair McCloskey from the University of Sheffield shares his experience using Reddit to engage wider audiences with research. Reddit offers a significant platform for social scientists to disseminate work and engage. Despite only occasionally submitting content to the site, the Faculty website received nearly ten times the amount of traffic from Reddit compared to its referrals from Twitter.

I am going to attempt a beginner’s guide to using Travis-CI for continuous integration for R packages. It is going to be a beginner’s guide because that is all I could possibly write; my knowledge and experience with Travis is limited. Sometimes it can be helpful to have someone walk you through something new that she herself has only recently come to grips with, though, so here we go!

… We need a way to recover for workflows the nice properties the ML platforms are already giving us for each step the workflow is made of.

In the sofware engineering field, we have long known a very effective method to cure the kind of problems outlined above. To increase the abstraction level of our workflow descriptions and free them from inessential detail, we must express our solution in a new language that is closer to the domain at hand, i.e., machine learning and feature engineering. In other words, we need a domain-specific language, or DSL, for machine learning workflows.

Consider TPOT your Data Science Assistant. TPOT is a Python tool that automatically creates and optimizes machine learning pipelines using genetic programming.