Data Science newsletter – May 19, 2021

Oxford University has partnered with global technology corporation Oracle, to create a Global Pathogen Analysis System (GPAS) that enables scientists and decision-makers to quickly identify COVID-19 variants.

GPAS combines Oxford University’s Scalable Pathogen Pipeline Platform (SP³) with Oracle Cloud Infrastructure (OCI). SP³, which was originally used to sequence tuberculosis, has been repurposed to unify, analyse and compare different SARS-CoV-2 sequence data, producing annotated genomic sequences and identifying new variants of COVID-19.

An in-built analytics dashboard simplifies this process by highlighting which strains of the virus are spreading most quickly and which genetic features contribute to increased transmissibility and vaccine resistance.

Spatial analysis of daily Covid-19 cases at the US county scale revealed a dynamic multifractal scaling of infections, spanning from 10 to 2,600 km and consistently trending toward that of the susceptible population. A susceptible–infected–recovered model was expanded to include spatial spread across counties using a spatial kernel. The reproduction number Rb (average number of persons infected by an infected person) decreased because of interventions (masks, social distancing). The model shows that reducing Rb in isolation is not sufficient to stem the spread of the disease and concomitant measures such as curfews and lockdowns may be needed. The Rb of 2.0 estimated here in July to October 2020 is large, hinting at super-spreaders and super-spreader events. [full text]

Following a two-year study, Mayo Clinic has introduced a new screening process for a type of heart disease, utilizing artificial intelligence for increased accuracy in diagnosing the condition.

Systolic low ejection fraction, when the heart is unable to contract strongly enough to pump at least 50% of blood from its chamber at each beat, can be difficult to discover during early stages, and the typical diagnostic tool, an echocardiogram (ECG), requires more time and is less likely to pick up on asymptomatic cases.

Mayo Rochester developed an artificial intelligence-enabled electrocardiogram screening tool, conducting a trial with 45 Mayo rural and urban clinics, including Mayo Clinic Health System in La Crosse and Onalaska, with 348 clinicians randomly divided into usual care and intervention groups.

With the help of next generation technology, passengers at William P. Hobby Airport HOU, will get an upgrade to their travel experience, which includes live journey and wait times for passenger and social distance monitoring.

“We want our passengers to feel empowered by the technology we implement throughout their travel experience,” Houston Airport Director IT Program Management Diego Parra said. “This technology not only helps passenger know what to expect at certain points in their journey, but it also provides us with valuable information to keep them safe. When we see a congested area that needs greater social distancing, we can respond.”

In March 2020, countries around the world began to shut down to slow the spread of the coronavirus. Although the measures were preventative and protective, they also offered scientists an opportunity to evaluate how this drastic change rippled throughout the environment. Within a few weeks of the initial lockdown, Spanish scientists recorded a significant reduction in air pollution. Jordi Díaz, part of the scientific staff at Geociencias Barcelona–Consejo Superior de Investigaciones Científicas, and his team used this unique period to create a baseline for seismic noise in urban areas.

In September 2019, seismologists established a network of 14 temporary seismic sensors along with 5 permanent sensors (all placed 2 to 3 kilometers apart from one another) across Barcelona to evaluate different methods to capture ambient noise. What they did not anticipate was that the network would monitor the level of seismic vibrations during this remarkable period in human history. The study offers insight into how seismic monitors could be used to monitor human movement to understand economic activity. The results are available in a special issue of the journal Solid Earth.

We’ve always had a soft spot for language at Google. Early on, we set out to translate the web. More recently, we’ve invented machine learning techniques that help us better grasp the intent of Search queries. Over time, our advances in these and other areas have made it easier and easier to organize and access the heaps of information conveyed by the written and spoken word.

But there’s always room for improvement. Language is remarkably nuanced and adaptable. It can be literal or figurative, flowery or plain, inventive or informational. That versatility makes language one of humanity’s greatest tools — and one of computer science’s most difficult puzzles.

LaMDA, our latest research breakthrough, adds pieces to one of the most tantalizing sections of that puzzle: conversation.

CDS PhD student Artie (Yiqiu) Shen, alongside Assistant Professor and Emerging Scholar of Computer Engineering at NYU Abu Dhabi Farah Shamont and NYU School of Medicine student Jamie Oliver, has co-authored the “Artificial Intelligence System Reduces False-Positive Findings in the Interpretation of Breast Ultrasound Exams”. Additional co-authors include CDS PhD student Nan Wu and CDS affiliated professor Krzysztof J. Geras (who supervised the project), as well as several other NYU members.

During the early months of the COVID-19 pandemic, Jay Van Bavel, a psychologist at New York University, wanted to identify the social factors that best predict a person’s support for public-health measures, such as physical distancing or closing restaurants. He had a handful of collaborators ready to collect survey data. But because the pandemic was going on everywhere, he wondered whether he could scale up the project. So he tried something he’d never done before.

He posted a description of the study on Twitter in April, with an invitation for other researchers to join. “Maybe I’ll get ten more people and some more data points,” he recalls thinking at the time. Instead, the response floored him. More than 200 scientists from 67 countries joined the effort. In the end, the researchers were able to collect data on more than 46,000 people. “It was a massive collaboration,” he says. The team showed how, on the whole, people who reported that national identity was important to them were more likely to support public-health policies1. The work is currently being peer reviewed.

For social scientists, the COVID-19 pandemic has presented a unique opportunity — a natural experiment that “cuts across all cultures and socio-economic groups”, says Andreas Olsson, a psychologist at the Karolinska Institute in Stockholm. Everyone is facing similar threats to their health and livelihoods, “so we can see how people respond differently to this depending on culture, social groups and individual differences”, he says. Researchers have been able to compare people’s behaviours before and after large policy changes, for example, or to study the flow of information and misinformation more easily.

In a study published in the journal Soft Matter, the researchers demonstrate a machine learning technique that measures the topological traits of cell clusters. They showed that the system can accurately categorize cell clusters and infer the motility and adhesion of the cells that comprise them.

“You can think of this as topology-informed machine learning,” said Dhananjay Bhaskar, a recent Ph.D. graduate who led the work. “The hope is that this can help us to avoid some of the pitfalls that affect the accuracy of machine learning algorithms.”

Businesses would be wise to pay attention to the A.I. plans that many governments worldwide publish. Savvy companies can use the information to decide how to best invest in A.I.-related initiatives.

Recently, the Brookings Institution think tank and members of Queensland University of Technology analyzed the A.I. plans of 34 countries in an effort to explore the variation among the different plans. The analysis builds upon a prior policy paper from the authors that determined which industries governments believe would be most likely to be transformed by the technology, among other findings.

The winner: health care, which appeared in 28 plans.

Naval Surface Warfare Center, Crane Division (NSWC Crane) and Indiana University (IU) were awarded $1.7 million to collaborate on Artificial Intelligence (AI) programming for rural students. The Office of the Secretary of Defense (OSD) funded the AI pilot program, which will begin this summer for middle school students.

NSWC Crane is teaming up with two IU schools, the School of Education and the Luddy School of Informatics, Computing, and Engineering. NSWC Crane and IU received the National Defense Education Program (NDEP) Award from OSD to pilot the program, “AI Goes Rural: Middle School Artificial Intelligence Education” to Indiana middle schools.

Computational physics, comp chemistry, comp biology, comp medicine… Even comp archeology. Realistic simulations, big data analysis, and ML everywhere

Covering developments in neuromorphic computing has been something of a piecemeal experience, as happens with all novel architectures. There are few companies with scalable devices and the research is often specific to one experimental architecture or use case. In other words, no total picture has emerged for neuromorphic yet, especially for the datacenter.

Perhaps in response to this lack of ecosystem-wide clarity, a wide vendor and research collaboration has put together a roadmap for neuromorphic computing that provides past, present, and future challenges for the various approaches and most important, provides insight into the range of applications these technologies might capture first—if at all.

What is most notable about the neuromorphic roadmap is the absence of datacenter-centric applications. Creating scalable neuromorphic systems has been on Intel’s own roadmap with Loihi, for instance. The low power and potential for high performance networking held early promise, especially as the first wave of AI chip startups hit the market five years ago. However, the broad base of contributors to the roadmap rarely mention devices or applications that can address scalable distributed computing. Not that this is a surprise, after all, this is still a nascent area. But if we take this wide survey of neuromorphic in 2021 at face value, the area has shifted to an edge/embedded story.

Cornell is partnering in a $36 million grant from the Toyota Research Institute (TRI) for its Accelerated Materials Design and Discovery (AMDD) collaborative university research program, which seeks to use artificial intelligence to discover new materials that could help achieve emissions-free driving.

Not everyone can be a master painter or sculptor, but everyone has the potential to appreciate art. The skills and training to appreciate data must be a core competency taught within our education system. Without such education and training, data illiteracy will persist, further growing the data science divide that will keep many in the dark, to their own and society’s detriment.

… Even among astronauts who are highly prescreened for their ability to handle stress, conflict is both frequent and inevitable. “It happens on every single space mission,” Amanda Ripley told me. “Every single one.”

Ripley is the author of the new book, High Conflict: Why We Get Trapped and How We Get Out. Her reporting spanned astronauts and gang members to married couples and politicians, so I invited her to join me for a recent episode of “How To!” , to talk about strategies for turning “high conflict” — the kind that devolves into “me versus you” or “us versus them” — into productive conflict.