Data Science newsletter – November 23, 2019

GROUP CURATION: N/A

I am excited to announce that IBM Research is the first founding corporate partner of the Stanford Institute for Human-Centered Artificial Intelligence (HAI). Building on decades of research collaboration across computer and materials science, IBM is committed to joining HAI to advance AI research, education, policy and practice that improve how we live, work, play and learn.

In this new role, IBM will work closely with fellow AI thought leaders, researchers and innovators at Stanford and through participation on the HAI Corporate Advisory Committee. IBM researchers will also work alongside Stanford researchers as part of the Visiting Scholars program.

Across departments and with other academic, industry and government leaders, IBM will work with HAI to collaborate to advance the state of the art in AI research and to understand the multi-dimensional impact of AI as it transforms economies, business, legal and political norms, societies and cultures. By developing this understanding, IBM and HAI will seek to provide insight that will enable society to shape the application of AI in ways that are inclusive and promote social good.

There are plenty of reasons Deutsche Telekom AG and Vodafone Group Plc make for uneasy bedfellows. But if Europe’s biggest telecommunications firms can overcome their differences, they would benefit from forging a strong alliance for one of their biggest cost centers: towers.

The structures on which mobile operators install their antennas have generated a flurry of dealmaking as valuations soar and European carriers sense an opportunity to reduce debt and costs. By some estimates, towers account for a third of total capital expenditures. Since July, more than $8 billion of deals have been announced in Europe.

The women Ashley and I spoke with had specific reasons to location share: Dropping a pin so a friend can find her on the beach. Checking to see that a person who isn’t returning texts has simply made it home safely and passed out. Feeling an extra pair of eyes on her while she’s on a first date with a stranger. Making sure someone is not lying when they announce that they’re “five minutes away!” Some of these cases are about safety, but all are about accepting accountability. If you’re important to me, you should be able to find me. I trust you, so please watch me all the time!

U.S. colleges are stepping up scrutiny of student applications, as universities deal with the fallout from a scandal that saw parents pay bribes and rig test scores to win places for their children.

As the admission season begins, Yale University, Bowdoin College and Pomona College are among those conducting spot checks or verifying some information on applications to find signs of cheating or embellishment, according to school officials.

Growing speculation that Robert Jackson Jr. might step down from the Securities and Exchange Commission is sparking angst among Democrats, who worry that his departure would give the agency’s Republicans even more power to push a pro-Wall Street agenda.

Now a group of alumni and students at New York University School of Law, where Jackson is currently on leave, are weighing in as well –- and urging him not to rush back to his teaching gig. The effort includes an online petition with more than 100 signatories and a Facebook group called “NYU Law: Let SEC Commissioner Robert Jackson Serve His Country.”

A lot of companies use focus groups and surveys to understand how people feel. Now, emotional AI technology can help businesses capture the emotional reactions in real time — by decoding facial expressions, analyzing voice patterns, monitoring eye movements, and measuring neurological immersion levels, for example. The ultimate outcome is a much better understanding of their customers — and even their employees.

The Risks of Bias in Emotional AI

Because of the subjective nature of emotions, emotional AI is especially prone to bias. For example, one study found that emotional analysis technology assigns more negative emotions to people of certain ethnicities than to others. Consider the ramifications in the workplace, where an algorithm consistently identifying an individual as exhibiting negative emotions might affect career progression.

Google’s head of philanthropy says the company is having “a lot of conversations” internally amid worries about the tech giant’s bottomless appetite for consumer data and how it uses its algorithms.

Vice President Jacqueline Fuller wouldn’t comment on specific data privacy controversies dogging Google lately, but says she shares other concerns many have about Big Tech. Cyberbullying. Hate speech amplified online. The impact of artificial intelligence on everything, from jobs to warfare.

“As a consumer myself, as part of the general public, as a mother, it’s very important to understand … what am I seeing, what are my children seeing,” she said in an interview with The Associated Press in Paris, where she announced new grant winners Tuesday for projects aimed at teaching digital skills to poor, immigrant, rural or elderly users.

Health care–related AI interventions in LMICs can be broadly divided into three application areas (see the figure). The first includes AI-powered low-cost tools running on smartphones or portable instruments. These mainly address common diseases and are operated by nonspecialist community health workers (CHWs) in off-site locations, including local centers and households. CHWs may use AI recommendations to triage patients and identify those requiring close follow-up. Applications include diagnosing skin cancer from photographic images and analyzing peripheral blood samples to diagnose malaria (7); more are expected given the emergence of pocket diagnostic hardware, including ultrasound probes and microscopes. With increasing smartphone penetration, patient-facing AI applications may guide lifestyle and nutrition, allow symptom self-assessment, and provide advice during pregnancy or recovery periods—ultimately allowing patients to take control of their health and reducing the burden on limited health systems.

Microsoft Corp. co-founder Bill Gates discussed protectionism in technological research around topics like artificial intelligence. Gates argued that open systems will inevitably win out over closed ones. He speaks with John Micklethwait at Bloomberg’s New Economy Forum in Beijing. [video, 3:40]

Following the raising of a $90 million Series C, privately held GenapSys will begin rolling out a line of low cost electronic-based genetic sequencing machines that it says could upend the research and clinical testing industries.

CEO Hesaam Esfandyarpour says the Redwood City, Calif.–based company’s Sequencer, a table-top device that carries a price tag that’s about 1/100th of existing systems, will make will make genetic sequencing available to the vast majority of research institutions that can’t currently afford larger systems that can cost up to $1 million.

Engineers at Duke University have developed a microscope that adapts its lighting angles, colors and patterns while teaching itself the optimal settings needed to complete a given diagnostic task.

In the initial proof-of-concept study, the microscope simultaneously developed a lighting pattern and classification system that allowed it to quickly identify red blood cells infected by the malaria parasite more accurately than trained physicians and other machine learning approaches.

In June of 2018, the Society for Public Health Education partnered with The George Washington University Milken Institute School of Public Health, University of Maryland School of Public Health at College Park, and the Public Good Projects to convene the inaugural Digital Health Promotion Executive Leadership Summit in Washington, D.C. We launched the 2018 Summit for several reasons. First, because of the ubiquity of digital technology and reach of social media, it had become apparent to us that the time was right to bring greater focus to the promise that digital technology and social media hold for improving the public’s health. Second, while other forums were discussing uses of mHealth and eHealth, it was clear the public and private sectors were talking within siloes; thus we believed that more needed to be done to catalyze a dialogue among the academic, industry, and government sectors to share their efforts and to examine the opportunities—and challenges—of forging collaborative partnerships in developing and evaluating the next generation of digital information and health communication technologies. And third, we hoped to facilitate a conversation among leaders from each of these sectors that would result in a consensus on a “Common Agenda” for future actions to advance digital health promotion efforts.

The U.S. National Institutes of Health last week released a draft policy that will require all investigators with NIH funding to make their data sets available to colleagues. For the first time, grantees holding any NIH-funded grant—not just those above a $500,000 threshold in direct costs—will need to submit a detailed plan for sharing data, including steps to protect the privacy of research subjects.

For the biomedical research community, the draft rules, which update a 2003 policy, aren’t a big surprise: NIH has been gathering input on its ideas for the past 3 years. “We knew there was a lot of anxiety in the community about this and so we slowly shared our thinking and its evolution,” says Carrie Wolinetz, associate director for science policy at NIH in Bethesda, Maryland. Now, she says, “This is what the policy looks like in draft form.”

NIH will collect comments through 10 January 2020 and hopes to finalize the new rules next year. Here is more background on the policy.

Last week saw two releases from organisations which will have a big impact on the next decade in academic research. Following on from the US Government’s OPEN data policy in January, the National Institute of Health (NIH) has released their draft open data policy! The NIH is seeking public input on a trans-NIH data management and sharing policy proposal that further advances the Agency’s commitment to responsible data management and sharing. Meanwhile on the other side of the world, CODATA released the The Beijing Declaration on Research Data – and it seems great minds think alike.

Google Arts & Culture Lab collaborated with Wayne McGregor to turn his archive into a creative tool, using machine learning. Now, anyone, anywhere can take inspiration from McGregor’s body of work to create their own piece of dance.

Living Archive uses U-MAP to map almost half a million moments of movement from Wayne McGregor’s 25-year archive, and organise them by visual similarity.

With work related stress costing companies billions worldwide, companies are turning to mindfulness to alleviate this problem. But most people who practice mindfulness or meditation do it outside of their working hours. What if there was a way to integrate these practices at work? Current Mindfulness is a Chrome extension that wants you to take mindful breaks during the day, right at your desk. I sat down with Ananya Huria, the Co-founder of Current to learn more about her life experiences with mindfulness and Current’s inception.

As scientists increasingly reach across disciplinary divides, the success of collaborations can hinge on how they choose to exchange knowledge. Our analysis suggests that their approach should be based on an understanding of the way that different teams initiate research and the subjects they choose to cover. We offer some practical tips for researchers to improve the effectiveness of their interdisciplinary collaborations.

We introduce a reinforcement learning framework for retail robo-advising. The robo-advisor does not know the investor’s risk preference, but learns it over time by observing her portfolio choices in different market environments. We develop an exploration-exploitation algorithm which trades off costly solicitations of portfolio choices by the investor with autonomous trading decisions based on stale estimates of investor’s risk aversion. We show that the algorithm’s value function converges to the optimal value function of an omniscient robo-advisor over a number of periods that is polynomial in the state and action space. By correcting for the investor’s mistakes, the robo-advisor may outperform a stand-alone investor, regardless of the investor’s opportunity cost for making portfolio decisions.

“To help people assess some of the implications and unintended consequences of Differential Privacy, Caliper® is providing several maps for public inspection. This map, created with Maptitude®, shows the change in population for every Congressional district after applying Differential Privacy.”