Data Science newsletter – October 4, 2016

Perhaps one of the simplest ways for robots to teach each other is to pool information about their successes and failures in the world. Humans and animals acquire many skills by direct trial-and-error learning. During this kind of ‘model-free’ learning — so called because there is no explicit model of the environment formed — they explore variations on their existing behavior and then reinforce and exploit the variations that give bigger rewards. In combination with deep neural networks, model-free algorithms have recently proved to be surprisingly effective and have been key to successes with the Atari video game system and playing Go. Having multiple robots allows us to experiment with sharing experiences to speed up this kind of direct learning in the real world.

Enforcement Gaps Interface (EGI) uses a computational algorithm to mine hundreds of commercial sites for ads potentially containing illegal wildlife and wildlife products.

“[With] a lot of the wildlife for sale online, [it] is so difficult to tell whether or not it’s illegal. The only thing you can do is trust what the retailer’s saying; we already know that’s not a great idea,” said co-creator Jennifer Jacquet, Assistant Professor of Environmental Studies at New York University (NYU). “We think we could give enforcement or retailers a best guess with high, medium, low certainty that the product for sale is potentially illegal.”

An interdisciplinary team of New York University researchers, led by Winslow Burleson, PhD, MSE, BA, an associate professor at NYU’s Rory Meyers College of Nursing, received a $2.9M grant from the National Science Foundation (NSF) to develop a well-integrated software/hardware instrument incorporating visual, audio, and physical (haptics, objects, real-time fabrication) components, known as the NYU Holodeck.

“Our goal is to create an immersive, collaborative, virtual and physical research environment with unparalleled tools for intellectual and creative output,” said Professor Burleson.

The master’s degree business is booming. College graduates looking for a leg up in the job market are flocking to one- and two-year programs that promise entry to lucrative careers. Top colleges are more than willing to provide them — for a price. Tuition for a 30-credit master’s in computer science from the University of Southern California runs $57,000. Syracuse, Johns Hopkins and Carnegie Mellon charge over $43,000 for the same degree.

But one highly ranked program, at Georgia Tech, has taken a very different approach. Its master’s in computer science costs less than one-eighth as much as its most expensive rival — if you learn online. And a new study by Harvard economists found that in creating the program, Georgia Tech may have discovered a whole new market for higher education, one that could change the way we think about the problem of college costs.

This review covers recent advances in the implementation of enabling chemistry technologies into the drug discovery process. Areas covered include parallel synthesis chemistry, high-throughput experimentation, automated synthesis and purification methods, flow chemistry methodology including photochemistry, electrochemistry, and the handling of “dangerous” reagents. Also featured are advances in the “computer-assisted drug design” area and the expanding application of novel mass spectrometry-based techniques to a wide range of drug discovery activities. [full text]

National Science Foundation (NSF) announced $10 million in “Big Data Spokes” awards to initiate research in specific areas identified, supported, and organized by the Big Data Regional Innovation Hubs (BD Hubs). $2.4 million in Big Data Spoke, Early­ Concept Grants for Exploratory Research (EAGER) and planning awards will connect the Midwest Big Data Hub (MBDH) and midwestern data scientists, to support digital agriculture; community-driven and sustainable neuroscience data infrastructure; improved sensor technologies; citizen scientists and real-time air quality monitoring; and new data-to-decision systems in hazards management by partnering data scientists with emergency personnel.

Project topics range from precision agriculture to personalized education. The data spokes reflect the unique priorities and capabilities of the four BD Hubs, which represent consortia from the Midwest, Northeast, South and West of the country.

“The BD Spokes advance the goals and regional priorities of each BD Hub, fusing the strengths of a range of institutions and investigators and applying them to problems that affect the communities and populations within their regions,” said Jim Kurose, assistant director of NSF’s Computer and Information Science and Engineering Directorate. “We are pleased to be making this substantial investment today to accelerate the nation’s big data R&D innovation ecosystem.”

The shift of government, the private sector and consumers of coming together to tackle the biggest problems in healthcare is starting to show, and the best indication is the innovation in digital health. But in order to do more, all parties must continue working together and providing opportunities in the form of policies, partnerships and information.

That was the takeaway from a speech by Karen DeSalvo, HHS’s Acting Assistant Secretary for Health, at Health 2.0.

“We’ve seen some historic changes in the last eight years,” said DeSalvo. “And I think sometimes we begin to take them for granted, but it’s been an incredible opportunity to really put people first and recognize that they are at the center of all of this.”

Conceived by Computer Science and Engineering Department Head, Professor Nasir Memon, the summer pilot of “A Bridge to NYU Tandon” — a computer science bridge for non-computer scientists — drew a diverse group of students. There was the Princeton psychology major who always loved computers but was sidelined by dreams of Olympic pole vaulting; the mechanical engineer who landed a plum job on Wall Street but still dreamed of developing video games; the economics and anthropology student who saw the chance to create online education opportunities in underdeveloped nations; and the music technologist who understood that deep computer science knowledge would open a wide variety of career options.

Melinda Gates has spent the past 16 years seeding social change through the world’s largest private philanthropy, the Bill and Melinda Gates Foundation. She and her husband have presided over an endowment that has grown to $39.6 billion?—?and they’ve added Warren Buffett as a third steward. The foundation funds initiatives that aim to enhance healthcare and reduce extreme poverty globally, as well as expand educational opportunities and access to information technology in the United States. But even with such a broad mandate, it can’t do everything.

So Gates is stepping out on her own. Though she’ll continue to work on the foundation, she’s building up a personal office to dedicate resources and attention to an issue of central personal importance: getting more women into tech?—?and helping them stay there.

The Center for Data Science is excited to announce a new PhD Program in Data Science. The program will prepare students to advance the state-of-the-art in data science research and prepare them for outstanding careers in academia or industry. Admitted students are guaranteed financial support in the form of tuition and a stipend in the fall and spring semesters for up to five years. This support allows students to focus on their research goals independently of any grants that may become available through their research advisors. [video, 4:28]

“The WikiText language modeling dataset is a collection of over 100 million tokens extracted from the set of verified Good and Featured articles on Wikipedia.”

“The bayesboot package implements a function bayesboot that performs the Bayesian bootstrap introduced by Donald B. Rubin (1981). The implementation can both handle summary statistics that works on a weighted version of the data or that works on a resampled data set.”

Google Cloud Machine Learning brings the power and flexibility of TensorFlow to the cloud. You can use its components to select and extract features from your data, train your machine learning models, and get predictions using the managed resources of Google Cloud Platform.

Jenny Bryan, developer of the google sheets R package, gave a talk at Use2015 about the package.

One of the things that got me most excited about the package was an example she gave in her talk of using the Google Sheets package for data collection at ultimate frisbee tournaments. One reason is that I used to play a little ultimate back in the day.

Another is that her idea is an amazing one for producing cool public health applications. One of the major issues with public health is being able to do distributed data collection cheaply, easily, and reproducibly. So I decided to write a little tutorial on how one could use Google Sheets and R to create a free distributed data collecton “app” for public health (or anything else really).