Articles tagged with Algorithms
Scientists at CiTIUS have developed a new fast support vector classifier (FSVC) that significantly improves data classification using Machine Learning techniques. The FSVC is much faster and operates with less memory than traditional approaches, making it suitable for large-scale classification problems.
A new study developed a wearable technology-based method to assess myoclonus symptoms in the home environment. The method, which measures electrical neuromuscular function and movement, correlates well with assessments performed by experienced physicians.
Researchers developed a novel evidence-based material recommender system that predicts high entropy alloy formation without data descriptors, overcomes data bias and poor availability. The method recommends an FeMnCoNi alloy as the most probable HEA and successfully synthesizes it, confirming its validity.
A team of computer scientists has developed an assembly selection process that balances representation and fairness in citizens' assemblies. By using a machine learning-based algorithm, the researchers ensure that all volunteers have an equal chance of being chosen, regardless of demographic quotas or education level.
Lehigh University researcher Roberto Palmieri aims to make RDMA technology even faster by revisiting a long-held theory. His goal is to enable all machines to interact with local memory, reducing latency and improving performance.
Researchers from The University of Tokyo Institute of Industrial Science have identified the origin of a phenomenon that occurs when rubber materials under stress rapidly break. Their simplified step-loading model replicates the non-monotonic mechanical behavior observed in these materials, shedding light on the velocity jump phenomenon.
A machine learning model predicts new quasicrystals with nearly 71% accuracy, identifying key factors such as electron concentration and chemical composition. The discovery sheds light on the stabilization mechanism of quasicrystals, offering a step towards innovative materials like semiconductor and superconducting quasicrystals.
The University of Washington is leading a new NSF institute focused on using artificial intelligence to understand dynamic systems, which describe chaotic situations where conditions are constantly shifting and hard to predict. The institute aims to integrate fundamental AI theory with applications in critical technological areas.
Researchers from Skoltech and their colleagues developed a neural network that can efficiently generate IUPAC names for organic compounds in accordance with the IUPAC nomenclature system. The network, trained using the Transformer architecture, achieved an accuracy of nearly 99%, outperforming traditional rule-based solutions.
A new algorithm has been developed that can guide a quadrotor drone through a series of waypoints on a circuit, beating the fastest lap of two world-class human pilots. The algorithm generates time-optimal trajectories that fully consider the drones' limitations, resulting in faster and more consistent flight times.
Researchers from Skoltech have developed a new augmentation technique called MixChannel to help train computer vision algorithms with limited data. This approach outperformed state-of-the-art solutions in testing with three neural networks and can be combined with other methods for even more training data.
A team of scientists from Michigan State University is using artificial intelligence to analyze plant genomes and predict the functions of unknown genes. With a $1.4 million NSF grant, they aim to help farmers grow crops that can withstand drought and disease.
Researchers have developed new algorithms to record and display color in digital images with greater realism. The methods improve color accuracy for electronic displays and create more natural LED lighting.
Researchers at Columbia University School of Engineering and Applied Science have developed a computer vision technique that enables machines to predict human behavior with higher accuracy. The algorithm leverages higher-level associations between people, animals, and objects to make more intuitive predictions about future actions, ope...
The University of Houston's Air Quality Forecasting and Modeling Lab has developed an artificial intelligence system that can accurately predict ozone levels up to two weeks in advance. This breakthrough could lead to improved ways to control high ozone problems and contribute to solutions for climate change issues.
A team of researchers developed an algorithm that can predict when narwhals hunt, opening up new insights into their life and behavior. The algorithm uses acoustic data from clicks and buzzing sounds to map the whales' movement patterns and identify hunting areas.
Researchers aim to improve search results by accounting for users' systematic biases, which can lead to misleading information and poor decision-making. The project seeks to build bias-aware prediction models and promote informed, unbiased decision-making through proactively capturing bias-related search problems.
Using a computer-aided, multi-material 3D-printing method, researchers have created custom healthcare devices with tailored shapes and functions to enhance patient wellbeing. The technique offers a solution for mass-produced medical devices by providing bespoke designs that can improve comfort and durability.
Researchers developed an AI algorithm using 20,000 nights of sleep data to improve diagnoses and treatments for sleep disorders. The algorithm can analyze patient sleep patterns in seconds, freeing up medical hours that can be used to assess and diagnose more patients.
A team of bioengineers defined ten overarching classes of constraints on early metabolic networks based on basic chemistry and physics. These abiotic constraints dictate fundamental aspects of chemical processes, influencing the evolution of alternative transport systems and stress response mechanisms.
Researchers at Stanford University have developed a new prototype software to facilitate pattern-making in foundation paper piecing quilting. The algorithm is based on a graph structure called a hypergraph and allows users to design patterns without having to consider the geometry, ordering, and constraints of the technique.
A new algorithm allows solving applied problems like electrodynamics and hydrodynamics with a significant time gain of up to 50%. The method uses asynchronous processes and non-overlapping subdomains, providing accurate results with less computation time.
Researchers developed algorithms to solve curvature-dependent image processing problems using convex optimization, inspired by Euler's elastic curves and Gestalt laws of perception. The new models perform as well as current deep-learning algorithms but provide a deeper understanding of the structures learned.
Researchers at Brown University have developed an algorithm that can remove electrical artifacts from brain signals, enabling simultaneous recording and stimulation in deep brain stimulation devices. This breakthrough could lead to more effective DBS therapy with fewer side effects.
A RUDN mathematician has developed an algorithm that allows for the optimal distribution of computing tasks between IoT devices and the cloud, resulting in a reduction of power and time costs by about three times. The algorithm also protects devices from malware during data transmission.
Researchers developed a neuromorphic chip that can detect high-frequency oscillations and recognize complex biosignals in real time. The technology has the potential to improve neurosurgical interventions and monitor epilepsy outside of hospitals.
A team of UK scientists aims to improve understanding of Deep Learning algorithms' decision-making process, making them more trustworthy. The project will combine theory, modeling, data, and computation to unlock the next generation of deep learning.
A new AI algorithm developed by TUM and Imperial College London can detect pneumonia in pediatric x-ray images while maintaining patient privacy. The algorithm uses federated learning, secure aggregation, and differential privacy to keep data onsite and prevent individual identification.
Scientists at JAIST explored the connection between game-playing search tree algorithms and human decision-making, proposing two indicators - PPN and SCN - to analyze potential future states. These findings have applications in AI, optimization problems, and entertainment analysis.
Researchers developed a machine learning technique that measures topological traits of cell clusters, accurately categorizing them and inferring cellular behavior. The algorithm uses persistent homology to examine microscope images and identifies persistent topological objects that can be used to classify cells.
Research shows that removing 'race correction' from pulmonary function test results reveals a higher prevalence and severity of lung disease among Black patients. The study found an increase in the percentage of patients with any pulmonary defect from 59.5% to 81.7%, indicating a significant difference.
A research team from Iowa State University has developed a way to control hard-to-predict systems. The technique uses quotienting to identify the least fixed-point operator, which can result in a new model that acts as a supervisor of the system.
A new computational method enables researchers to design functional magnonic devices in a shorter time, overcoming traditional limitations. The approach uses inverse design and intelligent algorithms to create devices with multiple functionalities.
Researchers have created a software that can decode brain signals to translate mental handwriting into text on a computer screen. The technology, called 'mindwriting,' has been tested on a man with full-body paralysis and achieved speeds of over 18 words per minute, rivaling those of able-bodied peers.
A new brain-computer interface (BCI) system has been developed to restore communication in people with spinal cord injuries and neurological disorders. The system uses mental handwriting to record brain activity, allowing participants to compose sentences and communicate at a speed comparable to typing on a smartphone.
A scientist from HSE University has developed an image recognition algorithm that speeds up real-time processing of video-based image recognition systems by up to 40%. The algorithm uses a sequence of convolutional layers and fine-tuning to achieve accurate results while controlling loss in accuracy.
Researchers developed AI algorithms that assess unconsciousness in patients based on brain activity with high accuracy and reliability. The algorithms outperformed existing methods in real-world testing, distinguishing between conscious and unconscious states in surgery patients.
Eliminating race-based adjustments in equations to estimate kidney function may lead to significant changes in medication prescribing practices for Black adults, according to a new study. The findings suggest that this change could result in lower eGFR values and potentially impact treatment outcomes for diabetic patients.
A KAIST research team creates T-GPS, enabling the processing of large-scale graphs without storing them in main memory or on disks. This allows for faster computation and reduced costs compared to traditional approaches.
A new real-time 3D motion tracking system combines transparent light detectors with advanced neural network methods to enable fast tracking speed, compact hardware, and lower cost compared to existing solutions. The technology has promising applications in automated manufacturing, biomedical imaging, and autonomous driving.
Researchers adapted an algorithm to identify similarities in escort ads, highlighting common parts and indicating potential human trafficking activity. The InfoShield algorithm outperformed other algorithms at identifying trafficking ads with 85% precision, flagging them without false positives.
Researchers developed an algorithm to screen for pulmonary arterial hypertension in scleroderma patients, outperforming standard methods. The algorithm correctly identified all 10 patients with the condition in a study of 68 subjects.
Researchers created synthetic hearts using statistical shape analysis to identify how heart shape relates to function. The study included 1000 new synthetic hearts made open access for testing and analysis.
A recent study by University of Georgia researchers found that people are more likely to rely on algorithms for decision-making, even when faced with complex challenges. As the task becomes harder, people prefer algorithm-generated suggestions over human advice or 'wisdom of the crowd', according to the study.
Researchers at Penn State develop a new framework for meta-assembly, allowing for accurate reconstruction of full-length mRNA sequences from short fragments. This enables better understanding of cell machinery and potential biomarkers for diagnosis.
Researchers at Shibaura Institute of Technology developed a camera-based support system that accurately detects tactile paving. The new image processing algorithm can detect tactile paving up to 91.65% of the time, improving safety for visually impaired individuals.
Researchers at UOC are developing a system to analyze and digitize agricultural data, using machine learning and intelligent algorithms to personalize customer targeting and improve sales processes. The project aims to give small enterprises more competitive tools, with potential impact on other sectors.
Christopher Musco, an assistant professor at NYU Tandon, received a $562,224 NSF grant to develop new algorithms for processing massive data sets. His project aims to democratize big-data modeling and advance US mathematics education through closer ties with STEM fields.
Researchers used ship movement data to estimate wave conditions in real time, without extra costs for measurement instruments. The approach combines cause-and-effect calculations with ship geometry and dynamic motions.
A new satellite-measured Solar-induced Chlorophyll Fluorescence (SIF) product is introduced to enhance carbon neutrality research. The SIF product, developed using a physical-based algorithm named IAPCAS/SIF, retrieves emission data from two micro-windows within the O2 A-band.
UofL and Medtronic will develop a closed-loop system to monitor and adjust stimulator settings for multiple functions, including locomotion and bladder control. The project aims to improve the lives of people with spinal cord injury by reducing manual interaction and increasing positive outcomes.
A machine learning algorithm has been shown to accurately predict adult patients at risk of attempting or experiencing suicidal thoughts. The Vanderbilt Suicide Attempt and Ideation Likelihood (VSAIL) model uses routine health records to identify those who should receive further screening, potentially directing limited resources where ...
After solving the sum of cubes puzzle for 42, researchers found a new solution for 3 using an optimized algorithm that utilized mathematical 'sieving' techniques. The discovery suggests there may be infinitely many more solutions for other values of k, with the next solution involving gigantic 21-digit numbers.
Clinical algorithms that adjust for race can cause harm to patients by delaying access to additional tests and treatments, according to a new article. The authors argue that the medical community must step up to achieve consensus on paths forward to address racial inequities in health care.
A multidisciplinary team of scientists has developed a new way to detect phase transitions in raw data by analyzing its intrinsic dimension, a statistical property that reveals collective properties of partition functions. This method is agnostic and does not require prior knowledge of the system's parameters.
A new study proposes a modeling framework for V2V and V2I cooperative driving, which prioritizes saving fuel and reducing emissions. The automated algorithm saves energy by 7% under light traffic and 23% under heavy traffic along the corridor.
A recent study published in Information & Management found that human analysis can effectively identify fake online reviews by leveraging linguistic characteristics. Researchers asked participants to distinguish between authentic and fake hotel reviews, with surprisingly effective outcomes.
Researchers at Princeton University have created a new microchip design that enables AI systems to process data closer to the edge of computing infrastructure. The innovative design reduces power consumption and latency, making AI more accessible and efficient.
Researchers at Texas A&M University have developed an algorithm that uses reinforcement learning to automate the prediction of subsurface environments. The algorithm accurately forecasts oil and gas reserves by providing feedback on rewards and penalizing unfavorable predictions.
A new AI algorithm enabled a fully automated artificial pancreas system to achieve improved glucose control without manual input of meal or exercise information. This breakthrough could lead to more efficient and hassle-free diabetes management.