Computers Math

You would not be surprised to see an elephant in the savanna or a plate in your kitchen. Based on your prior experiences and knowledge, you know that is where elephants and plates are often to be found. If you saw a mysterious object in your kitchen, how would you figure out what it was? You would rely on your expectations or prior knowledge. Should a computer approach the problem in the same way? The answer may surprise you. Cold Spring Harbor Laboratory Professor Partha Mitra described how he views problems like these in a “Perspective” in Nature Machine Intelligence. He hopes his insights will help researchers teach computers how to analyze complex systems more effectively.
Mitra thinks it helps to understand the nature of knowledge. Mathematically speaking, many data scientists try to create a model that can “fit an elephant,” or a set of complex data points. Mitra asks researchers to consider what philosophical framework would work best for a particular machine learning task:
“In philosophical terms, the idea is that there are these two extremes. One, you could say “rationalist,” and the other “empiricist” points of view. And really, it’s about the role of prior knowledge or prior assumptions.”
Rationalists versus empiricists
A rationalist views the world through the lens of prior knowledge. They expect a plate to be in a kitchen and an elephant in a savanna.
An empiricist analyzes the data exactly as it is presented. When they visit the savanna, they no more expect to see an elephant than they do a plate. More

When bringing technologies into the workplace, it pays to be realistic. Often, for instance, bringing new digital technology into an organization does not radically improve a firm’s operations. Despite high-level planning, a more frequent result is the messy process of frontline employees figuring out how they can get tech tools to help them to some degree.
That task can easily fall on overburdened workers who have to grapple with getting things done, but don’t always have much voice in an organization. So isn’t there a way to think systematically about implementing digital technology in the workplace?
MIT Professor Kate Kellogg thinks there is, and calls it “experimentalist governance of digital technology”: Let different parts of an organization experiment with the technology — and then centrally remove roadblocks to adopt the best practices that emerge, firm-wide.
“If you want to get value out of new digital technology, you need to allow local teams to adapt the technology to their setting,” says Kellogg, the David J. McGrath Jr. Professor of Management and Innovation at the MIT Sloan School of Management. “You also need to form a central group that’s tracking all these local experiments, and revising processes in response to problems and possibilities. If you just let everyone do everything locally, you’re going to see resistance to the technology, particularly among frontline employees.”
Kellogg’s perspective comes after she conducted an 18-month close ethnographic study of a teaching hospital, examining many facets of its daily workings — including things like the integration of technology into everyday medical practices.
Some of the insights from that organizational research now appear in a paper Kellogg has written, “Local Adaptation Without Work Intensification: Experimentalist Governance of Digital Technology for Mutually Beneficial Role Reconfiguration in Organizations,” recently published online in the journal Organization Science. More

First, it takes skill to fly the often expensive pieces of equipment smoothly and without crashing. And once you’ve mastered flying, there are camera angles, panning speeds, trajectories and flight paths to plan.
With all the sensors and processing power onboard a drone and embedded in its camera, there must be a better way to capture the perfect shot.
“Sometimes you just want to tell the drone to make an exciting video,” said Rogerio Bonatti, a Ph.D. candidate in Carnegie Mellon University’s Robotics Institute.
Bonatti was part of a team from CMU, the University of Sao Paulo and Facebook AI Research that developed a model that enables a drone to shoot a video based on a desired emotion or viewer reaction. The drone uses camera angles, speeds and flight paths to generate a video that could be exciting, calm, enjoyable or nerve-wracking — depending on what the filmmaker tells it.
The team presented their paper on the work at the 2021 International Conference on Robotics and Automation this month. More

North Carolina State University researchers have developed a computer simulation tool to predict when and where pests and diseases will attack crops or forests, and also test when to apply pesticides or other management strategies to contain them.
“It’s like having a bunch of different Earths to experiment on to test how something will work before spending the time, money and effort to do it,” said the study’s lead author Chris Jones, research scholar at North Carolina State University’s Center for Geospatial Analytics.
In the journal Frontiers in Ecology and the Environment, researchers reported on their efforts to develop and test the tool, which they called “PoPS,” for the Pest or Pathogen Spread Forecasting Platform. Working with the U.S. Department of Agriculture’s Animal and Plant Health Inspection Service, they created the tool to forecast any type of disease or pathogen, no matter the location.
Their computer modeling system works by combining information on climate conditions suitable for spread of a certain disease or pest with data on where cases have been recorded, the reproductive rate of the pathogen or pest and how it moves in the environment. Over time, the model improves as natural resource managers add data they gather from the field. This repeated feedback with new data helps the forecasting system get better at predicting future spread, the researchers said.
“We have a tool that can be put into the hands of a non-technical user to learn about disease dynamics and management, and how management decisions will affect spread in the future,” Jones said.
The tool is needed as state and federal agencies charged with controlling pests and crop diseases face an increasing number of threats to crops, trees and other important natural resources. These pests threaten food supplies and biodiversity in forests and ecosystems.
“The biggest problem is the sheer number of new pests and pathogens that are coming in,” Jones said. “State and federal agencies charged with managing them have an ever-decreasing budget to spend on an ever-increasing number of pests. They have to figure out how to spend that money as wisely as possible.”
Already, researchers have been using PoPS to track the spread of eight different emerging pests and diseases. In the study, they described honing the model to track sudden oak death, a disease that has killed millions of trees in California since the 1990s. A new, more aggressive strain of the disease has been detected in Oregon.
They are also improving the model to track spotted lanternfly, an invasive pest in the United States that primarily infests a certain invasive type of tree known as “tree of heaven.” Spotted lanternfly has been infesting fruit crops in Pennsylvania and neighboring states since 2014. It can attack grape, apple and cherry crops, as well as almonds and walnuts.
The researchers said that just as meteorologists incorporate data into models to forecast weather, ecological scientists are using data to improve forecasting of environmental events — including pest or pathogen spread.
“There’s a movement in ecology to forecast environmental conditions,” said Megan Skrip, a study co-author and science communicator at the Center for Geospatial Analytics. “If we can forecast the weather, can we forecast where there will be an algal bloom, or what species will be in certain areas at certain times? This paper is one of the first demonstrations of doing this for the spread of pests and pathogens.”
Story Source:
Materials provided by North Carolina State University. Original written by Laura Oleniacz. Note: Content may be edited for style and length. More

Stanford University computer science graduate student Mackenzie Leake has been quilting since age 10, but she never imagined the craft would be the focus of her doctoral dissertation. Included in that work is new prototype software that can facilitate pattern-making for a form of quilting called foundation paper piecing, which involves using a backing made of foundation paper to lay out and sew a quilted design.
Developing a foundation paper piece quilt pattern — which looks similar to a paint-by-numbers outline — is often non-intuitive. There are few formal guidelines for patterning and those that do exist are insufficient to assure a successful result.
“Quilting has this rich tradition and people make these very personal, cherished heirlooms but paper piece quilting often requires that people work from patterns that other people designed,” said Leake, who is a member of the lab of Maneesh Agrawala, the Forest Baskett Professor of Computer Science and director of the Brown Institute for Media Innovation at Stanford. “So, we wanted to produce a digital tool that lets people design the patterns that they want to design without having to think through all of the geometry, ordering and constraints.”
A paper describing this work is published and will be presented at the computer graphics conference SIGGRAPH 2021 in August.
Respecting the craft
In describing the allure of paper piece quilts, Leake cites the modern aesthetic and high level of control and precision. The seams of the quilt are sewn through the paper pattern and, as the seaming process proceeds, the individual pieces of fabric are flipped over to form the final design. All of this “sew and flip” action means the pattern must be produced in a careful order. More

So-called quantum dots are a new class of materials with many applications. Quantum dots are realized by tiny semiconductor crystals with dimensions in the nanometre range. The optical and electrical properties can be controlled through the size of these crystals. As QLEDs, they are already on the market in the latest generations of TV flat screens, where they ensure particularly brilliant and high-resolution colour reproduction. However, quantum dots are not only used as “dyes,” they are also used in solar cells or as semiconductor devices, right up to computational building blocks, the qubits, of a quantum computer.
Now, a team led by Dr. Annika Bande at HZB has extended the understanding of the interaction between several quantum dots with an atomistic view in a theoretical publication.
Annika Bande heads the “Theory of Electron Dynamics and Spectroscopy” group at HZB and is particularly interested in the origins of quantum physical phenomena. Although quantum dots are extremely tiny nanocrystals, they consist of thousands of atoms with, in turn, multiples of electrons. Even with supercomputers, the electronic structure of such a semiconductor crystal could hardly be calculated, emphasises the theoretical chemist, who recently completed her habilitation at Freie Universität. “But we are developing methods that describe the problem approximately,” Bande explains. “In this case, we worked with scaled-down quantum dot versions of only about a hundred atoms, which nonetheless feature the characteristic properties of real nanocrystals.”
With this approach, after a year and a half of development and in collaboration with Prof. Jean Christophe Tremblay from the CNRS-Université de Lorraine in Metz, we succeeded in simulating the interaction of two quantum dots, each made of hundreds of atoms, which exchange energy with each other. Specifically, we have investigated how these two quantum dots can absorb, exchange and permanently store the energy controlled by light. A first light pulse is used for excitation, while the second light pulse induces the storage.
In total, we investigated three different pairs of quantum dots to capture the effect of size and geometry. We calculated the electronic structure with highest precision and simulated the electronic motion in real time at femtosecond resolution (10-15 s).
The results are also very useful for experimental research and development in many fields of application, for example for the development of qubits or to support photocatalysis, to produce green hydrogen gas by sunlight. “We are constantly working on extending our models towards even more realistic descriptions of quantum dots,” says Bande, “e.g. to capture the influence of temperature and environment.”
Story Source:
Materials provided by Helmholtz-Zentrum Berlin für Materialien und Energie. Note: Content may be edited for style and length. More

Quantum computers with their promises of creating new materials and solving intractable mathematical problems are a dream of many physicists. Now, they are slowly approaching viable realizations in many laboratories all over the world. But there are still enormous challenges to master. A central one is the construction of stable quantum bits — the fundamental unit of quantum computation called qubit for short — that can be networked together.
In a study published in Nature Materials and led by Daniel Jirovec from the Katsaros group at IST Austria in close collaboration with researchers from the L-NESS Inter-university Centre in Como, Italy, scientists now have created a new and promising candidate system for reliable qubits.
Spinning Absence
The researchers created the qubit using the spin of so-called holes. Each hole is just the absence of an electron in a solid material. Amazingly, a missing negatively charged particle can physically be treated as if it were a positively charged particle. It can even move around in the solid when a neighboring electron fills the hole. Thus, effectively the hole described as positively charged particle is moving forward.
These holes even carry the quantum-mechanical property of spin and can interact if they come close to each other. “Our colleagues at L-NESS layered several different mixtures of silicon and germanium just a few nanometers thick on top of each other. That allows us to confine the holes to the germanium-rich layer in the middle,” Jirovec explains. “On top, we added tiny electrical wires — so-called gates — to control the movement of holes by applying voltage to them. The electrically positively charged holes react to the voltage and can be extremely precisely moved around within their layer.”
Using this nano-scale control, the scientists moved two holes close to each other to create a qubit out of their interacting spins. But to make this work, they needed to apply a magnetic field to the whole setup. Here, their innovative approach comes into play.
Linking Qubits
In their setup, Jirovec and his colleagues cannot only move holes around but also alter their properties. By engineering different hole properties, they created the qubit out of the two interacting hole spins using less than ten millitesla of magnetic field strength. This is a weak magnetic field compared to other similar qubit setups, which employ at least ten times stronger fields.
But why is that relevant? “By using our layered germanium setup we can reduce the required magnetic field strength and therefore allow the combination of our qubit with superconductors, usually inhibited by strong magnetic fields,” Jirovec says. Superconductors — materials without any electrical resistance — support the linking of several qubits due to their quantum-mechanical nature. This could enable scientists to build new kinds of quantum computers combining semiconductors and superconductors.
In addition to the new technical possibilities, these hole spin qubits look promising because of their processing speed. With up to one hundred million operations per second as well as their long lifetime of up to 150 microseconds they seem particularly viable for quantum computing. Usually, there is a tradeoff between these properties, but this new design brings both advantages together.
Story Source:
Materials provided by Institute of Science and Technology Austria. Note: Content may be edited for style and length. More

The impact of deploying Artificial Intelligence (AI) for radiation cancer therapy in a real-world clinical setting has been tested by Princess Margaret researchers in a unique study involving physicians and their patients.
A team of researchers directly compared physician evaluations of radiation treatments generated by an AI machine learning (ML) algorithm to conventional radiation treatments generated by humans.
They found that in the majority of the 100 patients studied, treatments generated using ML were deemed to be clinically acceptable for patient treatments by physicians.
Overall, 89% of ML-generated treatments were considered clinically acceptable for treatments, and 72% were selected over human-generated treatments in head-to-head comparisons to conventional human-generated treatments.
Moreover, the ML radiation treatment process was faster than the conventional human-driven process by 60%, reducing the overall time from 118 hours to 47 hours. In the long term this could represent a substantial cost savings through improved efficiency, while at the same time improving quality of clinical care, a rare win-win.
The study also has broader implications for AI in medicine. More