Data Science newsletter – July 28, 2017

GROUP CURATION: N/A

There is a telling example of this Trumpian impulse—the desire not to know—in a small D.O.E. program that goes by its acronym, ARPA-E. ARPA-E was conceived during the George W. Bush administration as an energy equivalent of DARPA—the Defense Department’s research-grant program that had funded the creation of G.P.S. and the Internet, among other things. Even in the D.O.E. budget the program was trivial—$300 million a year. It made small grants to researchers who had scientifically plausible, wildly creative ideas that might change the world. If you thought you could make water from sunlight, or genetically engineer some bug so that it eats electrons and craps oil, or create a building material that becomes cooler on the inside as it grows hotter on the outside, ARPA-E was your place. More to the point: your only place. At any given time in America there are lots of seriously smart people with bold ideas that might change life as we know it—it may be the most delightful distinguishing feature of our society. The idea behind ARPA-E was to find the best of these ideas that the free market had declined to finance and make sure they were given a chance. Competition for the grants has been fierce: only two out of every hundred are approved. The people who do the approving come from the energy industry and academia. They do brief tours of duty in government, then return to Intel and Harvard.

The man who ran the place when it opened was Arun Majumdar. He grew up in India, finished at the top of his engineering class, moved to the United States, and became a world-class materials scientist. He now teaches at Stanford University but could walk into any university in America and get a job. Invited to run ARPA-E, he took a leave from teaching, moved to Washington, D.C., and went to work for the D.O.E. “This country embraced me as one of her sons,” he said. “So when someone is calling me to serve, it is hard to say no.” His only demand was that he be allowed to set up the program in a small office down the street from the Department of Energy building. “The feng shui of D.O.E. is really bad,” he explained. [long read]

If Senate Republicans pass a version of their health reform bill, and if a current House GOP spending plan is enacted, it could spell double jeopardy for the Centers for Disease Control and Prevention.

The House Appropriations Committee earlier this month voted to approve a spending plan that would reduce the CDC’s budget by $198 million from 2017. But its bill relies on a fund Republicans are intent on repealing potentially as soon as the coming fiscal year — meaning the agency could take an even steeper cut.

UC Davis is receiving a nearly $3 million award from the National Science Foundation to train the next generation of marine scientists under a new paradigm that puts a focus on policy at the front end of research.

Beginning in fall 2018, the research traineeship, “Sustainable Oceans: From Policy to Science to Decisions,” will train scientists to generate their research questions by first asking what decisions are facing policymakers. The program’s creators envision a future where researchers, policymakers and educators seamlessly integrate natural and socioeconomic information, and improve the scientific basis of decision-making in managing coastal natural resources.

Individual cells take lineage commitment decisions in a way that is not necessarily uniform. We address this issue by characterising transcriptional changes in cord blood-derived CD34+ cells at the single-cell level and integrating data with cell division history and morphological changes determined by time-lapse microscopy. We show that major transcriptional changes leading to a multilineage-primed gene expression state occur very rapidly during the first cell cycle. One of the 2 stable lineage-primed patterns emerges gradually in each cell with variable timing. Some cells reach a stable morphology and molecular phenotype by the end of the first cell cycle and transmit it clonally. Others fluctuate between the 2 phenotypes over several cell cycles. Our analysis highlights the dynamic nature and variable timing of cell fate commitment in hematopoietic cells, links the gene expression pattern to cell morphology, and identifies a new category of cells with fluctuating phenotypic characteristics, demonstrating the complexity of the fate decision process (which is different from a simple binary switch between 2 options, as it is usually envisioned).

It looks like the two tech titans are arguing about AI’s impact on humanity. Really they’re protecting their personal brands.

For the last year, NatureSweet Farms has been testing a new monitoring system at one of its tomato farms in Arizona. The ten-camera system employs an artificial intelligence algorithm to tell human farmers when the crops are in danger from pests or disease.

Adrian Almeida, the company’s chief innovation officer, told CNN’s Future Tense that the system has already improved harvests by four percent. He expects that number could reach 20 percent once the system is in full swing.

NASA’s next exoplanet sleuth will have slightly out-of-focus cameras, but the space agency remains confident the Transiting Exoplanet Survey Satellite, or TESS, will still be able to discover distant planets.

The application of methods for developing trading strategies based on AI, both in short-term time frames and for longer-term investing, is gaining popularity and there are a few hedge funds that are very active in this field. However, broad acceptance of this new technology is slow due to various factors, the most important being that AI requires investment in new tools and human talent. The majority of funds use fundamental analysis because this is what managers learn in their MBA programs. There are not many hedge funds that rely solely on AI. Application of AI is growing at the retail level but the majority of traders still use methods that were proposed in mid twentieth century, including traditional technical analysis, because they are easy to learn and apply.

Note that AI and ML are not only used to develop trading strategies but also in other areas, for example in developing liquidity searching algos and suggesting portfolios to clients. Therefore, with AI applications gaining ground, the number of humans involved in trading and investment decisions decreases and this obviously affects markets and price action. It is early to speculate on the overall effects this new technology will have on the industry but it is possible that extensive use of AI will result in more efficient markets with lower volatility for extended periods of time followed by occasional volatility spikes due to regime changes. This is possible because the impact of subjective evaluation of information by humans will be minimized and with that the associated noise. But that remains to be seen in practice.

Ten undergraduates from colleges and universities across the country are spending the summer at WPI working alongside data scientists, conducting groundbreaking research that could someday impact critical facets of society from healthcare to mobility.

As part of a $356,000, three-year program funded by the National Science Foundation, the Data Science REU (Research Experience for Undergraduates) internship gives students hands-on experience with faculty-designed research, working alongside those professors and their graduate students.

Mayor Muriel Bowser formally launched The Lab @ DC, a first-of-its-kind in-house data science team in the DC Government. The Lab, which will use data and evidence to inform decision making across DC Government, is located within the City Administrator’s Office of Performance Management and operates under the leadership of Director David Yokum, a long-time advocate of data scientists in government.

In the mushrooming quantity of data generated by modern society, the University of Virginia sees an opportunity for accomplishing good. UVA’s Data Science Institute serves as a hub around which researchers can probe and analyze massive data sets to glean insight that could help solve thorny problems and improve the quality of life.

“Data science is touching everything we do,” said Philip E. Bourne, director of the Data Science Institute and Stephenson Chair of Data Science. “The ability to analyze a large body of relevant data has a positive benefit for society. The projects we support have the potential to impact people’s lives in real and precise ways.”

Here is sample of the many projects UVA’s Data Science Institute supports.

Data scientists remain in high demand. But these days, many types of professionals can boost their career prospects if they have at least a basic level of “data literacy,” says Jonathan Cornelissen.

That’s part of the reason why he says his online education company, DataCamp, has been growing over the past couple of years. DataCamp, which recently raised $4 million from investors, offers approximately 80 short training courses that teach data science skills to beginners and help experts sharpen their abilities.

“In almost every sector, almost every company is struggling to find not just hardcore data science talent, but even people who are data literate,” says Cornelissen, the company’s co-founder and CEO. (He’s pictured above in the center, with co-founders Dieter De Mesmaeker (left) and Martijn Theuwissen.)

The fields of neuroscience and artificial intelligence (AI) have a long and intertwined history. In more recent times, however, communication and collaboration between the two fields has become less commonplace. In this article, we argue that better understanding biological brains could play a vital role in building intelligent machines. We survey historical interactions between the AI and neuroscience fields and emphasize current advances in AI that have been inspired by the study of neural computation in humans and other animals. We conclude by highlighting shared themes that may be key for advancing future research in both fields. [full text]

Over the past 10,000 years human cultures have expanded from small groups of hunter-gatherers to colossal and complexly organized societies. The secrets to how and why this major cultural transition occurred have largely remained elusive. In an article published on July 24 by Russell Gray and Joseph Watts in PNAS they outline how advances in computational methods and large cross-cultural datasets are beginning to reveal the broad patterns and processes underlying our cultural histories.

Robust and actionable population health data at the city and neighborhood level is traditionally hard to come by, especially if community planners are looking for datasets presented in a slickly-packaged online dashboard free for public use.

But a small team of data analysts, graphical designers, and public health experts from NYU and the National Resource Networks are filling the need for accurate, standardized, and accessible local-level data with the City Health Dashboard, an interactive online tool that allows users to explore a variety of social, economic, clinical, and environmental factors in key geographic areas.

School is out, and at most universities that means final exams have given way to fun and sun. While the semester wrap-up at Georgetown University had its share of parties, students in Washington’s first-ever Hacking for Defense course prepared for perhaps their biggest challenge yet: reshaping the American military. As entrepreneur and Stanford University professor Steve Blank, the founder of the Hacking for Defense initiative, wrote on his blog introducing the movement: “We will not lose because we had the wrong technology. We will lose because we couldn’t adopt, adapt and deploy technology at speed and in sufficient quantities to overcome our enemies.”

Accepting that mission, these resourceful undergraduates spent 10 weeks painstakingly working to intimately understand and respond to the U.S. government’s toughest national security challenges. Using Silicon Valley innovation methods based on the premise that true innovation comes from grasping a customer’s problem better than they do, the students iterated threat-busting artificial intelligence algorithms, hunter drones, and augmented reality glasses, among other products.

Amazon has started a secret skunkworks lab dedicated to opportunities in health care, including new areas such as electronic medical records and telemedicine. Amazon has dubbed this stealth team 1492, which appears to be a reference to the year Columbus first landed in the Americas.

The stealth team, which is headquartered in Seattle, is focused on both hardware and software projects, according to two people familiar. Amazon has become increasingly interested in exploring new business in healthcare. For example, Amazon has another unit exploring selling pharmaceuticals, CNBC reported in May.

Molecular gas is the raw material which fuels star formation throughout the universe. Now, using the revolutionary Atacama Large Millimeter Array (ALMA) telescope, an international team of scientists has conducted one of the largest studies of molecular gas in distant galaxy clusters — rare conglomerations containing hundreds of galaxies, trillions of stars, and dark matter.

Scientists from the Spitzer Adaptation of the Red-sequence Cluster Survey (SpARCS) collaboration observed the galaxies within these distant clusters as they were when the universe was only 4 billion years old. They found that they harbor larger molecular gas reservoirs compared to galaxies in found in more typical isolated environments with fewer galaxy neighbors, known as field galaxies.

“We expected to find molecular gas deficiencies in these cluster galaxies compared to the field,” says lead author Allison Noble, a postdoc at the MIT Kavli Institute for Astrophysics and Space Research. “Galaxies in nearby clusters are dead, lacking star formation activity and with little to no molecular gas. In these distant clusters, we are instead detecting gas-rich galaxies, but their star formation rates are on par with field galaxies.”

Nasdaq, Inc. (Nasdaq:NDAQ), a leading provider of real-time market data and analytics to large institutions, financial professionals, and retail investors, announced the addition of four new data sets to the Nasdaq Analytics Hub. The Nasdaq Analytics Hub debuted in May 2017 to address the growing needs of the buy-side to gain access to potential new trading and investment signals derived from the growing number of structured and unstructured data sets. The addition of these four new data sets both broadens and deepens the high value content within the Nasdaq Analytics Hub, providing insights from corporate filings, combinations of fundamental and technical factors, and two different approaches for long and short investing.

New York, NY October 4-8. Mondo.NYC is a festival and global business summit of music and tech industry insiders and innovators, emerging artists and their fans. [$$$]

We’ve found that adding adaptive noise to the parameters of reinforcement learning algorithms frequently boosts performance. This exploration method is simple to implement and very rarely decreases performance, so it’s worth trying on any problem

“Please note a passing build status indicates all listed APIs are available since the last update. A failing build status indicates that 1 or more services may be unavailable at the moment.”

By taking Uber’s most popular open source Visualization frameworks, and aligning their websites, documentation and examples, and providing links between the websites, we can make our various frameworks significantly easier to discover and use for both new and experienced users.

“Gina Gotthilf, VP of Growth at language education platform Duolingo, is reluctant to spill one of her favorite A/B tests, but it’s too good not to share.”