Computers Math

Car manufacturing workers, long haul airline pilots, coal workers, shop assistants — many employees are forced to undertake the difficult and sometimes distressing challenge of finding a new occupation quickly due to technological and economic change, or crises such as the COVID-19 pandemic.
To make the job transition process easier, and increase the chances of success, researchers from the University of Technology Sydney (UTS) and UNSW Sydney have developed a machine learning-based method that can identify and recommend jobs with similar underlying skill sets to someone’s current occupation.
The system can also respond in real-time to changes in job demand and provide recommendations of the precise skills needed to transition to a new occupation.
Developed by Dr Nikolas Dawson and Dr Marian-Andrei Rizoiu from the UTS Data Science Institute and Professor Mary-Anne Williams, the Michael J Crouch Chair in Innovation at UNSW Business School, the system is based on findings from their new study, “Skill-driven Recommendations for Job Transition Pathways,” published in the international journal PLOS ONE.
What are the benefits of using AI to find a job?
Dr Dawson says while workplace change is inevitable, if we can make the job transition process easier and more efficient, there are significant productivity and equity benefits not only for individuals, but also for businesses and government. More

Scientists in Australia and in the USA have solved a long-standing mystery about the Sun that could help astronomers predict space weather and help us prepare for potentially devastating geomagnetic storms if they were to hit Earth.
The Sun’s internal magnetic field is directly responsible for space weather — streams of high-energy particles from the Sun that can be triggered by solar flares, sunspots or coronal mass ejections that produce geomagnetic storms. Yet it is unclear how these happen and it has been impossible to predict when these events will occur.
Now, a new study led by Dr Geoffrey Vasil from the School of Mathematics & Statistics at the University of Sydney could provide a strong theoretical framework to help improve our understanding of the Sun’s internal magnetic dynamo that helps drive near-Earth space weather.
The Sun is made up of several distinct regions. The convection zone is one of the most important — a 200,000-kilometre-deep ocean of super-hot rolling, turbulent fluid plasma taking up the outer 30 percent of the star’s diameter.
Existing solar theory suggests the largest swirls and eddies take up the convection zone, imagined as giant circular convection cells as pictured here by NASA.
However, these cells have never been found, a long-standing problem known as the ‘Convective Conundrum’. More

Building prediction models for high entropy alloys (HEAs) using material data is challenging, as datasets are often lacking or heavily biased. Now, researchers have developed a new evidence-based recommender system that determines various element combinations for potential HEAs. Unlike conventional data-driven techniques, this method has the added ability to recommend potential HEA candidates from limited amounts of experimental data. Their method can facilitate the development of alloys that have applications as advanced functional materials.
High entropy alloys (HEAs) have desirable physical and chemical properties such as a high tensile strength, and corrosion and oxidation resistance, which make them suitable for a wide range of applications. HEAs are a recent development and their synthesis methods are an area of active research. But before these alloys can be synthesized, it is necessary to predict the various element combinations that would result in an HEA, in order to expedite and reduce the cost of materials research. One of the methods of doing this is by the inductive approach.
The inductive method relies on theory-derived “descriptors” and parameters fitted from experimental data to represent an alloy of a particular element combination and predict their formation. Being data-dependent, this method is only as good as the data. However, experimental data regarding HEA formation is often biased. Additionally, different datasets might not be directly comparable for integration, making the inductive approach challenging and mathematically difficult.
These drawbacks have led researchers to develop a novel evidence-based material recommender system (ERS) that can predict the formation of HEA without the need for material descriptors. In a collaborative work published in Nature Computational Science, researchers from Japan Advanced Institute of Science and Technology (JAIST), National Institute for Materials Science, Japan, National Institute of Advanced Industrial Science and Technology, Japan, HPC SYSTEMS Inc., Japan, and Université de technologie de Compiègne, France introduced a method that rationally transforms materials data into evidence about similarities between material compositions, and combines this evidence to draw conclusions about the properties of new materials.
The research team consisted of Professor Hieu-Chi Dam from JAIST and his colleagues, Professor Van-Nam Huynh, Assistant Professor Duong-Nguyen Nguyen, and Minh-Quyet Ha, PhD student (JAIST); Dr. Takahiro Nagata, Dr. Toyohiro Chikyow, and Dr. Hiori Kino (National Institute for Materials Science, Japan); Dr. Takashi Miyake, (National Institute of Advanced Industrial Science and Technology, Japan); Dr. Viet-Cuong Nguyen (HPC SYSTEMS Inc., Japan); and Professor Thierry Denœux (Université de technologie de Compiègne, France).
Regarding their novel approach to this issue, Prof. Hieu-Chi Dam elaborates: “We developed a data-driven materials development system that uses the theory of evidence to collect reasonable evidence for the composition of potential materials from multiple data sources, i.e., clues that indicate the possibility of the existence of unknown compositions, and to propose the composition of new materials based on this evidence.” The basis of their method is as follows: elements in existing alloys are initially substituted with chemically similar counterparts. The newly substituted alloys are considered as candidates. Then, the collected evidence regarding the similarity between material composition is used to draw conclusions about these candidates. Finally, the newly substituted alloys are ranked to recommend a potential HEA.
The researchers used their method to recommend Fe-Co-based HEAs as these have potential applications in next-generation high power devices. Out of all possible combinations of elements, their method recommended an alloy consisting of iron, manganese, cobalt, and nickel (FeMnCoNi) as the most probable HEA. Using this information as a basis, the researchers successfully synthesized the Fe0.25Co0.25 Mn0.25Ni0.25 alloy, confirming the validity of their method.
The newly developed method is a breakthrough and paves the way forward to synthesize a wide variety of materials without the need for large and consistence datasets of material properties as Prof. Dam explains, “Instead of forcibly merging data from multiple datasets, our system rationally considers each dataset as a source of evidence and combines the evidence to reasonably draw the final conclusions for recommending HEA, where the uncertainty can be quantitatively evaluated.”
While furthering research on functional materials, the findings of Prof. Dam and his team are also a noteworthy contribution to the field of computational science and artificial intelligence as they allow the quantitative measurement of uncertainty in decision making in a data-driven manner. More

People with autism spectrum disorder have difficulty interpreting facial expressions.
Using a neural network model that reproduces the brain on a computer, a group of researchers based at Tohoku University have unraveled how this comes to be.
The journal Scientific Reports published the results on July 26, 2021.
“Humans recognize different emotions, such as sadness and anger by looking at facial expressions. Yet little is known about how we come to recognize different emotions based on the visual information of facial expressions,” said paper coauthor, Yuta Takahashi.
“It is also not clear what changes occur in this process that leads to people with autism spectrum disorder struggling to read facial expressions.”
The research group employed predictive processing theory to help understand more. According to this theory, the brain constantly predicts the next sensory stimulus and adapts when its prediction is wrong. Sensory information, such as facial expressions, helps reduce prediction error.
The artificial neural network model incorporated the predictive processing theory and reproduced the developmental process by learning to predict how parts of the face would move in videos of facial expression. After this, the clusters of emotions were self-organized into the neural network model’s higher level neuron space — without the model knowing which emotion the facial expression in the video corresponds to.
The model could generalize unknown facial expressions not given in the training, reproducing facial part movements and minimizing prediction errors.
Following this, the researchers conducted experiments and induced abnormalities in the neurons’ activities to investigate the effects on learning development and cognitive characteristics. In the model where heterogeneity of activity in neural population was reduced, the generalization ability also decreased; thus, the formation of emotional clusters in higher-level neurons was inhibited. This led to a tendency to fail in identifying the emotion of unknown facial expressions, a similar symptom of autism spectrum disorder.
According to Takahashi, the study clarified that predictive processing theory can explain emotion recognition from facial expressions using a neural network model.
“We hope to further our understanding of the process by which humans learn to recognize emotions and the cognitive characteristics of people with autism spectrum disorder,” added Takahashi. “The study will help advance developing appropriate intervention methods for people who find it difficult to identify emotions.”
Story Source:
Materials provided by Tohoku University. Note: Content may be edited for style and length. More

Facial recognition software can be used to spot a face in a crowd; but what if it could also predict where someone else was in the same crowd? While this may sound like science fiction, researchers from Japan have now shown that artificial intelligence can accomplish something very similar on a cellular level.
In a study published in Frontiers in Cell and Developmental Biology, researchers from Nara Institute of Science and Technology (NAIST) have revealed that a machine learning program can accurately predict the location of proteins related to actin, an important part of the cellular skeleton, based on the location of actin itself.
Actin plays a key role in providing shape and structure to cells, and during cell movement helps form lamellipodia, which are fan-shaped structures that cells use to “walk” forwards. Lamellipodia also contain a host of other proteins that bind to actin to help maintain the fan-like structure and keep the cells moving.
“While artificial intelligence has been used previously to predict the direction of cell migration based on a sequence of images, so far it has not been used to predict protein localization,” says lead author of the study, Shiro Suetsugu. This idea came in during the discussion with Yoshinobu Sato at the Data Science Center in NAIST. “We therefore sought to design a machine learning algorithm that can determine where proteins will appear in the cell based on their relationship with other proteins.”
To do this, the researchers trained an artificial intelligence system to predict where actin-associated proteins would be in the cell by showing it pictures of cells in which the proteins were labeled with fluorescent markers to show where they were located. Then, they gave the program pictures in which only actin was labeled and asked it to tell them where the associated proteins were.
“When we compared the predicted images to the actual images, there was a considerable degree of similarity,” states Suetsugu. “Our program accurately predicted the localization of three actin-associated proteins within lamellipodia; and, in the case of one of these proteins, in other structures within the cell.”
On the other hand, when the researchers asked the program to predict where tubulin, which is not directly related to actin, would be in the cell, the program did not perform nearly as well.
“Our findings suggest that machine learning can be used to accurately predict the location of functionally related proteins and describe the physical relationships between them,” says Suetsugu.
Given that lamellipodia are not always easy for non-experts to spot, the program developed in this study could be used to quickly and accurately identify these structures from cell images in the future. In addition, this approach could potentially be used as a sort of artificial cell staining method to avoid the limitations of current cell-staining methods.
Story Source:
Materials provided by Nara Institute of Science and Technology. Note: Content may be edited for style and length. More

A new Monash review throws the spotlight on recent research in heterostructures of topological insulators and magnetic materials.
In such heterostructures, the interesting interplay of magnetism and topology can give rise to new phenomena such as quantum anomalous Hall insulators, axion insulators and skyrmions. All of these are promising building blocks for future low-power electronics.
Provided suitable candidate materials are found, there is a possibility to realise these exotic states at room temperature and without any magnetic field, hence aiding FLEET’s search for future low-energy, beyond-CMOS electronics.
“Our aim was to investigate promising new methods of achieving the quantum Hall effect,” says the new study’s lead author, Dr Semonti Bhattacharyya at Monash University.
The quantum Hall effect (QHE) is a topological phenomenon that allows high-speed electrons to flow at a material’s edge, which is potentially useful for future low- energy electronics and spintronics.
“However, a severe bottleneck for this technology being useful is the fact that quantum Hall effect always requires high magnetic fields, which are not possible without either high energy use or cryogenic cooling.”
“There’s no point in developing ‘low energy’ electronics that consume more energy to make them work!” says Dr Bhattacharyya, who is a Research Fellow at FLEET, seeking new generation of low-energy electronics. More

Since receiving a $25 million grant in 2019 to become the first National Science Foundation (NSF) Quantum Foundry, UC Santa Barbara researchers affiliated with the foundry have been working to develop materials that can enable quantum information-based technologies for such applications as quantum computing, communications, sensing, and simulation.
They may have done it.
In a new paper, published in the journal Nature Materials, foundry co-director and UCSB materials professor Stephen Wilson, and multiple co-authors, including key collaborators at Princeton University, study a new material developed in the Quantum Foundry as a candidate superconductor — a material in which electrical resistance disappears and magnetic fields are expelled — that could be useful in future quantum computation.
A previous paper published by Wilson’s group in the journal Physical Review Letters and featured in Physics magazine described a new material, cesium vanadium antimonide (CsV3Sb5), that exhibits a surprising mixture of characteristics involving a self-organized patterning of charge intertwined with a superconducting state. The discovery was made by Elings Postdoctoral Fellow Brenden R. Ortiz. As it turns out, Wilson said, those characteristics are shared by a number of related materials, including RbV3Sb5 and KV3Sb5, the latter (a mixture of potassium, vanadium and antimony) being the subject of this most recent paper, titled “Discovery of unconventional chiral charge order in kagome superconductor KV3Sb5.”
Materials in this group of compounds, Wilson noted, “are predicted to host interesting charge density wave physics [that is, their electrons self-organize into a non-uniform pattern across the metal sites in the compound]. The peculiar nature of this self-organized patterning of electrons is the focus of the current work.”
This predicted charge density wave state and other exotic physics stem from the network of vanadium (V) ions inside these materials, which form a corner-sharing network of triangles known as a kagome lattice. KV3Sb5 was discovered to be a rare metal built from kagome lattice planes, one that also superconducts. Some of the material’s other characteristics led researchers to speculate that charges in it may form tiny loops of current that create local magnetic fields. More

A new mathematical model for predicting infectious disease outbreaks incorporates fear — both of disease and of vaccines — to better understand how pandemics can occur in multiple waves of infections, like those we are seeing with COVID-19. The “Triple Contagion” model of disease and fears, developed by researchers at NYU School of Global Public Health, is published in the Journal of The Royal Society Interface.
Human behaviors like social distancing (which suppresses spread) and vaccine refusal (which promotes it) have shaped the dynamics of epidemics for centuries. Yet, traditional epidemic models have overwhelmingly ignored human behavior and the fears that drive it.
“Emotions like fear can override rational behavior and prompt unconstructive behavioral change,” said Joshua Epstein, professor of epidemiology at NYU School of Global Public Health, founding director of the NYU Agent-Based Modeling Laboratory, and the study’s lead author. “Fear of a contagious disease can shift how susceptible individuals behave; they may take action to protect themselves, but abandon those actions prematurely as fear decays.”
For instance, the fear of catching a virus like SARS-CoV-2 can cause healthy people to self-isolate at home or wear masks, suppressing spread. But, because spread is reduced, the fear can evaporate — leading people to stop isolating or wearing masks too early, when there are still many infected people circulating. This pours fuel — in the form of susceptible people — onto the embers, and a new wave explodes.
Likewise, fear of COVID-19 has motivated millions of people to get vaccinated. But as vaccines suppress spread and with it the fear of disease, people may fear the vaccine more than they do the infection and forego vaccination, again producing disease resurgence.
For the first time, the “Triple Contagion” model couples these psychological dynamics to the disease dynamics, uncovering new behavioral mechanisms for pandemic persistence and successive waves of infection. More