Articles tagged with Algorithms
Researchers found that subjects' trust in human agents was strongly linked to their accuracy, whereas the value placed on computer algorithms depended on whether or not their predictions matched. The study's findings highlight the brain mechanisms underlying these kinds of evaluations and have implications for decision-making.
Researchers found that making movies with 'Oscar appeal' comes with both risks and rewards, including increased box office receipts but also a decrease in audience enjoyment. The study used data on nearly 3,000 Oscar-eligible films to build an algorithm predicting Oscar nominations based on film characteristics.
UCLA researchers Gabriel Rossman and Oliver Schilke analyzed 25 years of data on mainstream cinema to discover the logic behind Hollywood's decisions. They found that filmmakers who make movies with "Oscar appeal" (often mocked as "Oscar bait") face a risk-and-reward structure similar to lobbyists, with costly bids for potential awards.
Researchers at MIT developed a new algorithm that can dramatically reduce the number of operations needed to solve the max-flow problem, making it possible to tackle even huge networks. The algorithm divides each graph into clusters of well-connected nodes and focuses on bottlenecks, allowing for efficient routing.
Researchers developed an algorithm that subtly modifies facial features to make a face more memorable. The system uses machine learning to analyze subtle trends in facial features and applies these changes to produce a new image. In tests, the algorithm succeeded in making faces more or less memorable in around 75% of cases.
Researchers have demonstrated that powerful probabilistic reasoning algorithms can be implemented using chemical reactions, enabling the creation of intelligent machines at tiny scales. This breakthrough could lead to the development of 'smart drugs' that can automatically detect, diagnose, and treat diseases.
Researchers create algorithm using group pictures to determine urban tribes, achieving 48% accuracy rate. The algorithm analyzes images as the sum of their parts and attributes to pick up social cues.
MIT researchers have developed an algorithm that can perform Fourier transforms using close to the theoretical minimum number of samples. This could significantly reduce the time it takes for medical devices like MRI machines to scan patients, or allow astronomers to take more detailed images of the universe.
Researchers at UC San Diego developed an algorithm that uses group pictures to determine urban tribes, achieving 48% accuracy. The algorithm analyzes facial features, clothing, and other attributes in a 'parts and attributes' approach, with potential applications in search results, social networks, and public space surveillance.
Computer scientists at Princeton University have developed a method that allows researchers to identify quickly the connections between seemingly disparate groups in large collections of information. By using a mathematical method to calculate the likelihood of a pattern repeating throughout a subset of data, the tool can cut dramatica...
Janice Beitz, a Rutgers-Camden nursing professor, has developed an ostomy algorithm to guide nurses in proper assessment and management of ostomy patients. The algorithm provides a step-by-step pathway to selecting the best ostomy management option, enhancing patient safety.
A new image analysis and segmentation algorithm helps quantify grain sizes in Martian rock and soil images, revealing subtle trends with composition. The semi-automated algorithm provides better consistency across multiple images than human segmentation.
Researchers have developed an algorithmic approach to analyzing breast cancer screening data, enabling more precise detection of malignant tumors. The machine learning method reduces false positives and unnecessary interventions, improving patient outcomes.
A recent study by MIT economists provides justification for current rules in job-placement programs, which can accommodate married couples. The researchers found that a properly calibrated system can work well even with large job markets and specific constraints. In fact, data from the medical job market suggests that stable matching i...
A new algorithm proposed by MIT professor Berthold Horn aims to alleviate traffic flow instabilities, which can cause unexplained traffic jams. The algorithm uses sensor information from cars behind and in front of each other to stabilize traffic flow and reduce the risk of jams.
A new algorithm based on the Parsons code successfully identifies individual dolphin whistles, reducing time and errors in classification. This innovation enables researchers to study dolphin social behavior and develop a better understanding of their communication.
Yedioth Group won the Daniel H. Wagner Prize for its analysis that reduced print magazine delivery costs by using sophisticated analytics and math modeling. The team developed a pooling model of delayed distribution, enabling dramatic reductions in production and return levels while maintaining sales.
Researchers have developed a method to accurately measure heart and respiration rates remotely, without physical contact, using a simple video camera paired with complex algorithms. The approach has been tested on humans, mice, pigs, and zebrafish, with high accuracy and reliability.
A new robot-vision algorithm based on the Bingham distribution can identify object orientations more accurately than existing algorithms, with a 50% improvement in cases of poor visual information. The algorithm combines probabilities from different sources and has been successfully tested in cluttered scenes.
A team of engineers has developed a new approach that marries computer vision and hardware optimization to sort cells up to 38 times faster than current techniques. The approach improves on imaging flow cytometry, which uses camera mounted on a microscope to capture cell features.
Researchers develop an algorithm for image stitching that minimizes seam artifacts by smoothing the transition between images, producing visually appealing results. The approach is based on minimizing an energy function and can be used for both two-dimensional and future three-dimensional image stitching applications.
Researchers at Scripps Research Institute have developed a new method, called Extensive Combinatorial Refinement (ExCoR), that combines existing formulas to create more accurate computer models of molecules. This process can help identify the best algorithms for refining structural details and improve the development of drug candidates.
A team of researchers has developed a program called RealBrush that allows graphic artists to quickly produce realistic brushstrokes on their computers. The program uses machine-learning approaches and Big Data storage techniques to create, bend, and shape various types of brushstrokes.
Researchers developed algorithm to improve tumor segmentation in medical images, reducing errors and devastating consequences.
Researchers at MIT have developed a software system, Jigsaw, that optimizes cache management for multicore chips, improving performance by an average of 18% and reducing energy consumption by up to 72%. The system adapts to changing computational patterns, allowing it to dynamically allocate cache space and minimize energy usage.
Researchers at the University of Miami discovered a global ecology of interacting robots that trade on global markets at speeds too fast for humans. These ultrafast algorithms create extreme events, such as crashes and spikes, in financial markets.
Researchers developed a new program to simulate protein movements by exploiting similarities with robot arms, enabling faster and cheaper analysis. The project combines mechanical engineering and biosciences, aiming to understand protein movement and its potential applications in diseases.
Researchers have developed a tri-linear interpolation algorithm for accurate white matter fiber tracking, resulting in longer tracked fibers with smoother orientations. This method achieved reliable tracking results that aligned with actual brain anatomy and disease conditions.
Researchers have identified a promising avenue for early diagnosis of Alzheimer's disease through non-invasive brain wave analysis. The study found statistically significant differences in brain wave patterns between healthy individuals and those with mild to moderate Alzheimer's disease.
A new software solution developed by TUM's Prof. Gritzmann and his team consolidates agricultural land more efficiently than ever before, saving farmers up to 30% on running costs. The algorithm uses sophisticated mathematical ideas to optimize field allocation and reduce travel costs.
Researchers at MIT and Max Planck Institute have mapped the wiring diagram of a tiny patch of mouse retina using a combination of human and artificial intelligence. They identified 950 neurons, including a new type of retinal cell, and classified most of them based on their connections with other neurons.
Researchers developed a new framework to analyze ad hoc networks with fluctuating link quality, providing mathematical bounds on message propagation efficiency. The framework describes algorithms that can achieve maximal efficiency using randomness and adversarial relationships.
Researchers developed a computer algorithm that can infer the rules of complex systems by analyzing intermittent samples. By looking at relative changes in population updates, they uncovered an invariant geometry that relates to the hidden reactions driving the system.
Researchers at MIT have developed a computer system, Remy, that automatically generates efficient TCP congestion-control algorithms. These algorithms significantly outperform human-designed versions in simulations, yielding transmission rates 2-3 times higher and delay reductions of up to two-thirds.
Scientists are facing a digital bottleneck in assembling genetic data, which is being produced at an unprecedented rate. Computing solutions are needed to address this issue, including clever algorithms and harnessing more computing power. The article highlights the need for innovative solutions to overcome this challenge.
Stanley Osher, a professor at UCLA, is awarded the 2013 John von Neumann Lecture for his groundbreaking work on non-oscillatory methods, level set methods, and l1 and TVD methods. His research has numerous applications in engineering, physics, and image processing.
A team from EPFL has developed a computer algorithm that can create a 3D model of a simple room based on sound picked up by four microphones. The algorithm uses the differences in signal time to calculate distances between microphones, walls, and the sound source.
Researchers at Carnegie Mellon University developed a method for tracking individuals in complex nursing home settings using video cameras. The technique used multiple cues such as apparel color, person detection, trajectory and facial recognition to track 13 people with an accuracy of 88%, outperforming leading algorithms.
A new, fast and accurate algorithm developed by Polish researchers can calculate the Chandrasekhar function with accuracy up to over a dozen decimal digits. This method is crucial for understanding physical and chemical properties of materials' surfaces studied under laboratory conditions.
Researchers from North Carolina State University have developed a software algorithm that can detect compromised agents in distributed network control systems, allowing for isolation and protection of the entire system. This provides resilience and security advantages over traditional centralized systems.
Researchers have developed a computer algorithm that can model and catalogue the entire set of lightweight, carbon-containing molecules that chemists could feasibly create in a lab. The map helps scientists identify unexplored regions of the chemical space where new compounds may hold solutions to some of the world's most vexing challe...
Researchers from North Carolina State University have developed a computer program to help evaluate and rank scientific articles in their field. The algorithm assigns relevancy scores to thousands of papers, allowing biocurators to focus on the most informative papers, increasing productivity by 27% and novel data content by 100%. Whil...
The new journal focuses on research advances in uncertainty quantification, covering topics like finance, disaster preparedness, and porous media flows. The inaugural volume features papers on novel methods for reducing computational complexity and estimating information content in data assimilation.
A large-scale evaluation of computational protein function prediction reveals that algorithms combining disparate prediction clues provide more accurate predictions, turning genomic data into useful information. The study provides a comprehensive picture of how well current methods perform, with potential improvements identified.
MIT researchers have developed an algorithm that solves graph Laplacians in nearly linear time, using a spanning tree to simplify the calculation. This approach has significant implications for various applications, including scheduling, image processing, and online product recommendation.
Researchers developed a pixelated vision system to enhance retinal implants, allowing patients to detect motion, large objects, and improve orientation. The system uses mathematical algorithms to pick out salient locations in images and provides directional cues for navigation tasks.
USC scientists have developed an algorithm to predict the value of DNA sequencing, which could make it more affordable for clinics. The algorithm uses a model from ecology known as capture-recapture to estimate population sizes, with potential applications in public health, astronomy, and biology.
A UPNA researcher has developed 3D reconstruction algorithms that are on par with the best available techniques, but can be executed more rapidly on a computer. These algorithms use stereoscopic vision to capture depth and can be implemented in real-time using standard graphics cards.
A new algorithm for message dissemination in decentralized networks is faster and deterministic, guaranteeing delivery to all nodes. The algorithm's efficiency relies on the collective action of nodes performing simple local tasks, emerging into a global behavior.
A systematic review of TB interventions in low-and-middle income countries found limited real-world evidence to support their effectiveness and financial value. The authors identified substantial gaps in published evidence for scale-up of these interventions, which may prevent program-wide implementation.
Researchers have developed a computer algorithm that accurately predicts lower back pain in patients 76% of the time using brain scans. The study suggests structural changes in gray matter are associated with chronic lower back pain and may reflect pathology within the brain.
Researchers have developed a groundbreaking mind-controlled prosthetic hand that allows users to control it with remarkable accuracy and naturalism. The device has been tested in a clinical trial, achieving a success rate of up to 91.6% and demonstrating clinically significant improvement.
Researchers at Arizona State University propose a novel approach to the origins of life, focusing on the 'software' - information content. The study suggests that life is characterized by its unique use of information, providing a roadmap for identifying criteria for emergence. This approach moves away from chemical-based approaches, w...
According to UCSB researchers, looking just below the eyes provides the best view of a person's face for identifying their identity, emotions and genders. The brain uses sophisticated computations to plan eye movements for accuracy in tasks like flight or love at first sight.
A team of researchers from the Public University of Navarre has developed a new algorithm for improving the real-time capture of 3-D images, enabling applications in fields like 3-D video recording and intelligent systems. The technique uses stereoscopic vision to compare pairs of images and yields better results than previous algorithms.
Researchers at Stanford University have developed a new algorithm called ReFIT that greatly improves the speed and accuracy of thought-controlled computer cursors. The system, which was tested on rhesus monkeys, can control the cursor with speeds approaching those of real arms, while previous systems saw decline in performance over time.
Researchers aim to characterize RNA molecule variation in human neurons and heart cells, which may provide insights into aging and disease. They will use novel technologies to analyze individual cell transcriptomes and explore the role of G protein-couple receptors.
Researchers at EPFL's Computer Graphics and Geometry Laboratory have created an algorithm to control the 'caustic' effect, a natural optical phenomenon that generates clear images on transparent surfaces. The technique allows for the creation of complex representations such as faces or landscapes from simple forms like stars.
Researchers from PNNL develop new algorithm to identify similar patterns in data, reducing processing time by seconds. A software called Bamboo is also presented, which translates MPI code to speed up data-crunching and reduce communication bottlenecks.
A research team led by Professor Mike Stilman at Georgia Institute of Technology hopes to develop a robot that can use objects in its environment to accomplish high-level tasks. The project aims to create a robot capable of using everyday objects as tools, similar to MacGyver, to overcome complex challenges.