Welcome to the November 19, 2021 edition of ACM TechNews, providing timely information for IT professionals three times a week.
ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
|
2021 ACM Gordon Bell Prize Awarded to Team Achieving Real-Time Simulation of Random Quantum Circuit
ACM
November 18, 2021
ACM named a team of Chinese researchers to receive the 2021 ACM Gordon Bell Prize for simulating a random quantum circuit (RQC) in real time using a Sunway supercomputer. The researchers developed a systematic design process encompassing the algorithm, parallelization, and architecture needed to model a 10x10x random quantum circuit (RQC), which realized 1.20-Eflops single-precision or 4.4-Eflops mixed-precision performance on more than 41.9 million Sunway cores. The project outperformed cutting-edge RQC simulation processes and generated the model in 304 seconds; the Summit supercomputer would have taken approximately 10,000 years to perform the same task using Google's Sycamore quantum processor. The researchers said they assumed this challenge to help in both quantum-device development and in bringing algorithmic and architectural advances to the supercomputing community.
|
UMaryland Medical System Drops Race-Based Algorithm
The Washington Post
Ovetta Wiggins
November 17, 2021
The University of Maryland Medical System (UMMS) and the University of Maryland School of Medicine have stopped using a race-based algorithm for diagnosing kidney function. The equation relies on concentrations of the muscle-protein metabolism byproduct creatinine, and accounts for age, gender, and whether a patient is "African American or non-African American." The American Society of Nephrology's Susan E. Quaggin said the race modifier's inclusion creates the impression of better kidney function in Black patients, who often qualify for treatment or are listed for a kidney transplant later than they should as a result. UMMS' Stephen Seliger is collaborating with others to implement a new diagnostic equation that eliminates race in January.
*May Require Paid Registration
|
Software Uses Selfies to Detect Early Symptoms of Parkinson's Disease
University of Rochester NewsCenter
November 15, 2021
Computer vision software based on algorithms developed by University of Rochester researchers analyzes smartphone videos, including clips created while taking selfies, to predict an individual’s likelihood of developing Parkinson's disease. The software detects subtle movements of facial muscles that do not register on the naked eye. The test requires individuals to perform various facial movements during a short video, so the algorithm can calculate the likelihood that person is exhibiting symptoms of Parkinson's or related disorders. When patients smile, the software can detect "modularity," a Parkinson's symptom indicating a loss of control over facial muscles.
|
Programming Languages: Why Former Favorite Is Sliding Down the Rankings
ZDNet
Liam Tung
November 12, 2021
PHP’s popularity declined in software-testing firm Tiobe’s November list of the most popular progamming languages, falling two places to 10th place compared with last November’s ranking. Tiobe CEO Paul Jansen said PHP, which is used frequently for Web and backend development, faces significant rivalry, but since many small and medium enterprises still rely on the scripting language, "I expect PHP to decline further but in a very slow pace." PHP remains popular on GitHub, but has been overtaken by Microsoft's JavaScript superset language Typescript, which features a type system that compiles into JavaScript and is favored for larger Web applications.
|
Improved Computational Model to Analyze Antibody Patterns
UCLA Samueli Newsroom
November 17, 2021
University of California, Los Angeles (UCLA) researchers have designed an advanced computational model to analyze human antibody levels by streamlining datasets on antibodies' molecular virus-binding interactions. The researchers used the tensor decomposition model to construct a matrix that simulated disease progression before mining the data to identify otherwise hard-to-find antibody patterns, and applied this technique to two separate antibody datasets involving COVID-19 and HIV patients. The results yielded insights about the interconnected relationships between the profiled antibodies and the immune system response. UCLA's Aaron Meyer said, "This study shows how such antibody patterns can be greatly simplified and, in turn, help in the design of better therapies."
|
AI System to Help Fight Spread of Infections in Hospitals
The Pittsburgh Post-Gazette
November 17, 2021
University of Pittsburgh and Carnegie Mellon University (CMU) scientists have developed an artificial intelligence (AI)-based technique for detecting and halting hospital-based infectious disease outbreaks. The researchers integrated machine learning and whole genome sequencing to enable identification of outbreaks much faster than with traditional methods. The AI system employs genomic sequencing surveillance to detect whether patients in hospitals have near-identical strains of an infection; computers then can mine patients' electronic health records to determine common transmission vectors, such as procedures using the same equipment or shared healthcare providers. The system “can quickly detect and characterize an emerging outbreak, helping hospitals take swift and precise actions to stop a bug from spreading," said CMU's Artur Dubrawski.
|
Universe Simulation Spanning Billions of Light-Years is Largest Ever
ScienceAlert
Matt Williams
November 17, 2021
Researchers at the Flatiron Institute's Center for Computational Astrophysics and the Harvard-Smithsonian Center for Astrophysics say the AbacusSummit simulation suite they created is the largest cosmological model ever produced. Comprised of over 160 N-body simulations, AbacusSummit can process nearly 60 trillion particles to model their gravity-influenced behavior in a box-shaped environment. N-body simulations are essential to modeling dark matter's interaction with baryonic or visible matter, and the researchers ran the models on the Summit supercomputer at the U.S. Department of Energy's Oak Ridge Leadership Computing Facility. The researchers designed the Abacus codebase to update 70 million particles per node/second; the codebase can analyze the simulation while it is running to search for patches of dark matter that signal the presence of bright star-forming galaxies.
|
Robots Can Use Their Own Whirring to Echolocate, Avoid Collisions
New Scientist
Matthew Sparkes
November 18, 2021
Robots can navigate and avoid collisions using the sounds they produce through echolocation as bats do, according to researchers at Denmark's Aalborg University and France's Universite de Lorraine. Aalborg's Jesper Rindom Jensen and colleagues proposed robots can detect obstacles like walls or other robots by picking up sounds reflected off those objects. Onboard computers can measure the time it takes for noise from the robot to reach a surface and be reflected back to a microphone on the robot, detecting obstacles as far off as one meter (three feet). Earlier research yielded a device that beamed sound around itself to navigate, but laboratory experiments demonstrated that background noise created by the robot can accomplish the same task.
|
What Went Wrong With Zillow? A Real Estate Algorithm Derailed Its Big Bet
The Wall Street Journal
Will Parker; Konrad Putzier
November 17, 2021
Real estate firm Zillow Group had looked to its digital home-flipping business Zillow Offers to lead its growth in the future, but the company has acknowledged that will not happen because the unit’s underlying algorithm could not reliably predict housing prices. That failure, the company said, was rooted in the technology's inability to understand the real estate market and predict housing prices, which are shaped by fluctuating factors like aesthetics and regional factors that influence buyers' decisions. Zillow CEO Richard Barton admitted to shareholders that the algorithm could not accurately predict swings in home prices, and the company is closing Zillow Offers.
*May Require Paid Registration
|
Can a Machine Learn Morality?
The New York Times
Cade Metz
November 19, 2021
Morality is a thorny issue for machines, as scientists learned in testing Delphi, a system programmed by the Allen Institute for Artificial Intelligence (AI) to make moral judgments. The neural network analyzed more than 1.7 million ethical judgments made by humans to establish a morality baseline for itself, and people generally agreed with its decisions when it was released to the open Internet. Some, however, have found Delphi to be inconsistent, illogical, and insulting, highlighting how AI systems reflect the bias, arbitrariness, and worldview of their creators. Delphi's developers hope to build a universally applicable ethical framework for AI, but as Zeerak Talat at Canada's Simon Fraser University observed, "We can't make machines liable for actions. They are not unguided. There are always people directing them and using them."
*May Require Paid Registration
|
AI Recognizes Primate Behaviors in the Wild
University of Oxford (U.K.)
November 17, 2021
Researchers and wildlife conservationists will be able to streamline analysis of animal behavior in video footage through a new artificial intelligence model that can identify chimpanzees' behavioral patterns in the wild. A team of scientists from the U.K., Japan, and the U.S. trained the model on videos from two chimpanzee populations in West Africa to automatically recognize wild primate behavior. Said Max Bain at the U.K.'s University of Oxford, "We use methods from deep learning with networks that are able to ingest both the audio and the visual stream of information from a video, crucial in the wild where an animal behavior might be heard but not seen, or inaudible actions." The technique combines individual identification from previous audio-visual behavior recognition methods to investigate animal behaviors more thoroughly.
|
|
