NYU Data Science newsletter – April 20, 2016

GROUP CURATION: N/A

The Healthcare Information Management Systems Society (HIMSS) organizes the largest health information technology conference in North America every year. This year’s event, held in Las Vegas, attracted about 42,000 attendees from around the world. … Following are four key conference takeaways and trends to keep an eye on:

1. Precision medicine is close to reality.

Once upon a time, about 13.8 billion years ago, our universe sprang from a quantum speck, ballooning to one million trillion trillion trillion trillion trillion trillion times its initial volume (by some estimates) in less than a billionth of a trillionth of a trillionth of a second. It then continued to expand at a mellower rate, in accordance with the known laws of physics.

So goes the story of cosmic inflation, the modern version of the Big Bang theory. That single short, outrageous growth spurt fits all existing cosmological data well and accounts for the universe’s largeness, smoothness, flatness and lack of preferred direction. But as an explanation of how and why the universe began, inflation falls short. The questions it raises — why the growth spurt happened, how it happened, what (if anything) occurred beforehand — have confounded cosmologists since the theory emerged in the 1980s. “We have very strong evidence that there was this period of inflation,” said Matthew Kleban, a cosmologist at New York University. “But we have no idea — or we have many, many ideas — too many ideas — what inflation was, fundamentally.”

Elodie Fichet, a Communication doctoral candidate studying public relations and crisis communication, was featured in several articles in the past couple of weeks surrounding her research with Assistant Professor Kate Starbird on the impact of online rumoring on Twitter following crisis events.

AOL and the Jacobs Technion-Cornell Institute at Cornell Tech announced in a press statement they had come up with something called immersive recommendations technology. This would let users opt in to use their personal digital information from one platform to better inform content recommendations for them on another platform. … This is the first technology to emerge from the Connected Experiences Lab collaboration between Cornell Tech and AOL, which aims to develop ideas in digital technology.

While hackathons — events where tech-savvy people collaborate to develop new projects — are getting more common, this Saturday’s Flawless Hacks hackathon had one unique feature: all the participants were women.

During the event, engineers from Make School, Airbnb, and Shopify led workshops on iOS development, web development and version control. In addition, mentors from companies such as Teachers Pay Teachers, The New York Times, Google, Dropbox, and Facebook worked with various teams on their projects.

Organized by NYU students, Flawless Hacks targets women entering STEM fields at all levels of expertise. CAS senior Kira Prentice and Steinhardt senior Kaitlin Gu started the project in February.

That was the topic discussed recently by OCLC Research Library Partners Research metadata managers, initiated by John Riemer of UCLA. With increasing expectations that research data creation made possible through grant funding will be archived and made available to others, many institutions are becoming aware of the need to collect and curate this new scholarly resource. To maximize the chances that metadata for research data are shareable (that is, sufficiently comparable) and helpful to those considering re-using the data, our communities would benefit from sharing ideas and discussing plans to meet emerging discovery needs.

There’s a replication crisis in biomedicine—and no one even knows how deep it runs.

The use of technology, no matter how advanced, does not guarantee competitive advantage. For technology to advance business strategy, it must be embedded in what we call the “integrated strategy machine.”

As the social and economic benefits of transit become clearer and clearer, a parade of data-driven maps and websites have tried to evaluate transit access in major American cities: where buses and trains go, who they serve, how effectively, and how often.

Tuesday marks the launch of AllTransit, the most exhaustive and accessible such resource yet. A joint project of the Center for Neighborhood Technology and TransitCenter, it assembles the largest collection of transit data anywhere—543,000 transit stops, 800 transit agencies, and 15,000 routes nationwide, according to the site. That in itself is a major public service, since agencies aren’t (as of yet) required by the DOT to open up their data about connectivity, access, and frequency. AllTransit doesn’t offer that data raw (not for free, at least), but it does offer a number of useful ways to explore it.

Crop diseases are a major threat to food security, but their rapid identification remains difficult in many parts of the world due to the lack of the necessary infrastructure. The combination of increasing global smartphone penetration and recent advances in computer vision made possible by deep learning has paved the way for smartphone-assisted disease diagnosis. Using a public dataset of 54,306 images of diseased and healthy plant leaves collected under controlled conditions, we train a deep convolutional neural network to identify 14 crop species and 26 diseases (or absence thereof). The trained model achieves an accuracy of 99.35% on a held-out test set, demonstrating the feasibility of this approach. When testing the model on a set of images collected from trusted online sources – i.e. taken under conditions different from the images used for training – the model still achieves an accuracy of 31.4%. While this accuracy is much higher than the one based on random selection (2.6%), a more diverse set of training data is needed to improve the general accuracy. Overall, the approach of training deep learning models on increasingly large and publicly available image datasets presents a clear path towards smartphone-assisted crop disease diagnosis on a massive global scale.

An ambitious plan to create the world’s first centre of excellence and research building for the study of collective animal behaviour is now a reality with a multi-million euro investment into Iain Couzin’s cross-disciplinary program at the University of Konstanz.

The Centre for Visual Computing of Collectives (VCC) will provide a world-leading facility for the collaboration of computer scientists and behavioural biologists to address some of society’s biggest challenges—from insect plagues to disease spread to robotic intelligence

For fifty years, the Hawaii International Conference on System Sciences (HICSS) has been a home for researchers in the information, computer, and system sciences (http://www.hicss.org/). … With an eye to the exponential growth of digitalization and information networks in all aspects of human activity, HICSS has continued to expand its track on Digital and Social Media (http://www.hicss.org/#!track3/c1xcj).

Waikola Village, Hawaii Wednesday-Saturday, January 4-7.

Deadline for submissions is Wednesday, June 15.

Boosting is a machine learning meta-algorithm that aims to iteratively build an ensemble of weak learners, in an attempt to generate a strong overall model.

Lets look at the highlighted parts one-by-one.

This is a (necessarily incomplete) list of women active in machine learning maintained by the Women in Machine Learning (WiML) organization. (See this link for more information about the annual WiML Workshop event.) If you are organizing an event related to machine learning, this list serves as a resource for potential speakers.

Whether it’s reading more books, learning a new language or working out regularly, achieving your goals can be really hard. One day it’s “I got called into a last-minute meeting.” The next day it’s “I have a friend in town.” And before you know it, your goals are delayed or forgotten. In fact, with all the things you need to do in a given week, it’s probably harder than ever to find the time—even when your goal really matters to you. That’s why starting today, we’re introducing Goals in Google Calendar. Just add a personal goal—like “run 3 times a week”—and Calendar will help you find the time and stick to it. [video, 1:23]