NYU Data Science newsletter – July 4, 2016

GROUP CURATION: N/A

In this episode of the O’Reilly Data Show, I spoke with Rajat Monga, who serves as a director of engineering at Google and manages the TensorFlow engineering team. We talked about how he ended up working on deep learning, the current state of TensorFlow, and the applications of deep learning to products at Google and other companies. [audio, 38:49]

On the latest episode of Recode Decode, hosted by Kara Swisher, the co-founders of Numenta (who previously co-founded Palm and Handspring) predicted AI will indeed drive technological advances for the rest of the century … but they threw water on the idea that intelligent machines would behave as they do in popular fiction. [audio, 43:30]

A multimillion-dollar funding drive is being readied to transform arXiv, the vastly popular repository to which physicists, computer scientists and math­ematicians flock to share their research preprints openly.

But the results of an enormous user survey published this week suggest that researchers are wary of drastic changes to a site that has become an essential part of the infrastructure of modern science.

London’s favored status among startups seeking a European base is in peril, and its counterparts on the continent are wasting no time in announcing themselves as worthy alternatives. Besides Dublin, there’s Berlin, already home to successes including Rocket Internet and Soundcloud; Amsterdam, which has spawned the major payments player Adyen; and Stockholm, with Spotify and Minecraft’s maker, Mojang.

Most academics are not naturally angry people. We live in a world where statements are made only after careful hypothesis testing, where data are analysed meticulously, where conclusions report what the data show. We are trained to never run ahead of the data, never prove the null hypothesis.

So, how do you make this group angry? You could start by attacking what matters to us most – say, the future of science and higher education. Then you could attack the people we care about – say, our international community of colleagues, research collaborators, friends – we see the contribution these people make to advancing disciplinary knowledge, teaching and supervising the leaders of the future. Then, you could attack the one thing we all defend to the hilt – the truth.

As the U.K.’s vote to leave the European Union shows, the act of one nation could have widespread implications around the world. Similarly, the U.S. housing market meltdown of 2007 to 2009 led to a global financial crisis the effects of which are still being felt today. These examples underscore the need by nations to pay attention not only to what goes on within their borders, but also to factors that lead to systemic risk that could bring down outside economies.

Research universities are critical contributors to our national research enterprise. They are the principal source of a world-class labor force and fundamental discoveries that enhance our lives and the lives of others around the world. These institutions help to create an educated citizenry capable of making informed and crucial choices as participants in a democratic society. However many are concerned that the unintended cumulative effect of federal regulations undercuts the productivity of the research enterprise and diminishes the return on the federal investment in research.

An analysis of Australian Research Council data reveals grant proposals that integrate a broad array of academic fields are less likely to be funded.

Opinion evolution and judgment revision are mediated through social influence. Based on a large crowdsourced in vitro experiment (n = 861), it is shown how a consensus model can be used to predict opinion evolution in online collective behaviour. It is the first time the predictive power of a quantitative model of opinion dynamics is tested against a real dataset. Unlike previous research on the topic, the model was validated on data which did not serve to calibrate it. This avoids to favor more complex models over more simple ones and prevents overfitting. The model is parametrized by the influenceability of each individual, a factor representing to what extent individuals incorporate external judgments. The prediction accuracy depends on prior knowledge on the participants’ past behaviour. Several situations reflecting data availability are compared. When the data is scarce, the data from previous participants is used to predict how a new participant will behave. Judgment revision includes unpredictable variations which limit the potential for prediction. A first measure of unpredictability is proposed. The measure is based on a specific control experiment. More than two thirds of the prediction errors are found to occur due to unpredictability of the human judgment revision process rather than to model imperfection.

An influential federal panel has taken the unusual step of telling the Obama administration to withdraw a controversial proposal to revise regulations that protect people who volunteer for medical research.

The proposal is “marred by omissions, the absence of essential elements, and a lack of clarity,” according to the National Academies of Sciences, Engineering and Medicine. The conclusions are part of a 283-page report released Wednesday.

The regulations are known collectively as the Common Rule. They were put in place decades ago to make sure medical experiments are conducted ethically.

The National Socio-Environmental Synthesis Center (SESYNC) located in Annapolis, Maryland, seeks proposals for independent, collaborative team-based synthesis research by graduate student teams. … The goal of this program is to engage graduate students in team-based socio-environmental synthesis research and transdisciplinary collaboration.

Deadline for proposals is Friday, August 26. We encourage those interested in these opportunities to discuss their ideas with SESYNC before Friday, July 29.

People who need access to data for their jobs are spending more and more time searching for data of interest to the task at hand. This is particularly true for data-driven companies, where the heterogeneity of data sources and storage systems is broader than ever before and where volumes of data are ever-increasing. We are building a data discovery system to solve this challenge, increasing the productivity and quality of data-based jobs. The system has an interface similar to a search engine, which allows anyone—regardless of technical skill level—to find data of interest. The interface accepts keywords and special tokens to assist users in discovering relevant data.

Communication is a two-way street. So as a first step, if you’re a scientist struggling with how to collaborate with journalists, recognize your role in communicating effectively.