Floating air particles following disasters and other geological events can have a lasting impact on life on Earth, and a new model drawing on chaos theory looks to help predict how these particles move, with an eye toward applications for geoengineering. Tímea Haszpra developed a model for following particles as they travel around the globe. Using it, she has generated maps that can be used to predict how particles will be dispersed above the world.
A deep reinforcement learning algorithm developed by computer scientists at the University of California, Irvine can solve the Rubik's Cube puzzle in a fraction of a second. The work is a step toward making AI systems that can think, reason, plan and make decisions.
Princeton researchers have built an electronic array on a microchip that simulates particle interactions in a hyperbolic plane, a geometric surface in which space curves away from itself at every point.
First national study shows cutting residents' training hours has not resulted in lower performance for new doctors. Resident training was capped at 80 hours per week in 2003, down from 100+ hours, a controversial move that left many worried. Despite worries, reduced hours did not change 30-day patient mortality, readmissions or spending.
Perfumes that use the most popular scents do not always obtain the highest number of ratings, according to an analysis of online perfume reviews.
In the contract-based demand response, some of the participants may default in providing the scheduled negawatt energy owing to demand-side fluctuations faults. Thus, the detection of defaulting participants is an important function of the aggregator. A group of Japanese researchers has developed a method to detect defaulting participants based on sparse reconstruction. This enables assured detection of defaulting participants with limited information that aggregator can utilize.
Instead of the typical bell-shaped curve, the fossil record shows a fat-tailed distribution, with extreme, outlier, events occurring with higher-than-expected probability. Using the same mathematical tools that describe stock market crashes, Santa Fe Institute scientists explain the evolutionary dynamics that give rise to universal patterns in the fossil record.
A team of scientists from the Universities of Oxford, Cornell and San Jose State, collaborating across theoretical and experimental physics and computer science, have developed and trained a new Machine Learning (ML) technique, to finally understand how electrons behave in important quantum materials. Their far-reaching results were published in Nature online on 19 June and will feature in this week's print issue of Nature (Thursday 27 June).
The function of protein machines in biological cells is so complex that even supercomputers cannot predict their cycles at atomic detail. But, as demonstrated in this review article, many aspects of their operation at mesoscales can be already revealed by exploring simple mechanical models, amenable for simulations on common computers. The authors also show how artificial protein-like structures with machine properties can be designed.
A new mathematical model of the structure of networks could help find new cancer drugs, speed up traffic flow and combat sexually transmitted disease. Although the three challenges seem diverse, they all could benefit from a theory that helps uncover information about a network by analyzing its structure. Successful link prediction algorithms already exist for certain types of networks, but the researchers analyzed differently structured networks to come up with their alternative algorithm.