Articles tagged with Algorithms
Researchers at UCLA have designed a mathematical algorithm to analyze crime patterns and identify street gangs involved in unsolved violent crimes. The algorithm, which was tested on real data from an East Los Angeles police district, correctly identified the most likely gang rivalries 80% of the time.
Dr. Modesto Orozco has been awarded an ERC Advanced Grant for his project SimDNA, a computational biology initiative aiming to unravel the structure and properties of nucleic acids. The five-year grant provides funding of around 2-3 million Euros.
Researchers have developed an algorithm that recognizes linguistic cues to divide texts into probable author groupings, sidestepping methodological hurdles in Bible scholarship. The software was able to separate passages from two Hebrew books with 99 percent accuracy and distinguish between 'priestly' and 'non-priestly' materials.
University of Texas at Austin researchers have discovered an optimal algorithm to extract information from individual images to determine focus error. This breakthrough could improve auto-focusing in digital cameras and deepen understanding of human depth perception.
Researchers at UT Knoxville have developed an algorithm that improves the accuracy of electrocardiograms (ECGs) using smartphone technology. The algorithm can detect electrode misplacement and electromagnetic noise, providing more accurate A-F letter grades for ECGs and recommendations for optimal electrode placement.
A new algorithm developed by Tel Aviv University researchers can process massive amounts of digital data instantly, allowing for more efficient tracking and prediction of criminal movements. The algorithm creates probability maps displaying likely locations of suspects, enabling officials to focus resources on high-probability areas
A new algorithm using GPS data from a tracking system can predict and prevent collisions between small aircraft. Researchers at MIT have developed the algorithm to address the challenge of false alarms while allowing for some error margin.
Researchers at MIT's Computer Graphics Group have developed new methods to efficiently simulate blur in digital computer graphics, reducing rendering time by minutes. By simplifying the computation of blur, these techniques can create more convincing video games and frames of digital video, improving overall visual quality.
The study compares three popular ranking methods, finding that the Colley and Massey methods are insensitive to small changes, while the Markov method is highly sensitive, leading to anomalies in rankings. The authors conclude that these methods can be effective for other applications, such as social networks.
A new study from MIT neuroscientists shows that the most memorable photos are those containing people, followed by static indoor scenes and human-scale objects. The researchers developed a computer algorithm to rank images based on memorability, which could be useful for graphic designers and photo editors.
The grant enables scientists to monitor the world's oceans and track changes in coastal or maritime systems, supporting a more sophisticated understanding of global climate change. Dr. Stamnes' algorithms will also aid in exploring issues related to polarization, scattering, and refraction of light in atmosphere-ocean systems.
A new algorithm allows for high-capacity data hiding in executable files, providing a secure way to embed and conceal data. The technology can be used on various devices, including smartphones and tablets.
A recent study found that teenagers retain a unique and consistent pattern of underlying brain oscillations even during periods of rapid neural change, lending support to the idea of a 'brain fingerprint',
Julien Langou receives $400,000 NSF CAREER award to study algorithms and their limitations in numerical simulations. His research aims to better understand how fast the fastest algorithm can execute on a given computing platform.
Researchers at Carnegie Mellon University have discovered a biologically-inspired method to organize wireless sensor networks and distributed computing applications. Inspired by the fruit fly's nervous system, a new algorithm has been developed that can efficiently deploy sensors without advance knowledge of their connections.
A new national program using Carnegie Mellon University's computer algorithm successfully matched transplant candidates with living donors, enabling the first kidney transplants. The program aims to increase the number of kidney paired-donation transplants by creating a national pool of mismatched donor-recipient pairs.
A national pilot program utilizing a Carnegie Mellon-developed algorithm is matching living kidney donors with medically compatible transplant candidates. The initial run of the computer matching process successfully matched seven people in need of a kidney transplant with suitable donors, expanding kidney paired-donation transplants.
A study found significant variation in medical center performance of traditional central line-associated BSI surveillance. The researchers discovered that different institutions use varying criteria for surveillance, leading to inconsistent rates and rankings.
Researchers enabled by AFOSR's YIP program developed state-of-the-art detection and evasion algorithms to block known and future attacks on Air Force computer networks and wireless systems. The new technology will enhance network security with high accuracy and speed.
A new algorithm by Carnegie Mellon researchers can efficiently solve systems of linear equations, a crucial problem in computer applications like image processing and logistics. The algorithm applies to symmetric diagonally dominant (SDD) systems, outperforming existing methods by billion-fold speedup.
The ICIAM Collatz Prize recognizes individual scientists under 42 worldwide for outstanding research in industrial and applied mathematics, while the Lagrange Prize honors career-long contributions. Other notable prizes include the Maxwell, Pioneer, and Su Buchin Prizes.
Researchers will model complex neural systems to understand how humans see, then embed parallel transactions into silicon systems. A series of prototypes with breadboard hardware will be created every six months as understandings emerge.
Researchers developed new algorithms to simulate global mantle flow, plate tectonics, and individual fault zones, achieving a resolution of about one kilometer near the plate boundaries. The model accurately predicted plate movements, including anomalous rapid motion of microplates in the western Pacific.
The 2010 International Mathematical Union prizes were awarded to four mathematicians: Elon Lindenstrauss, Ngô Bảo Châu, Stanislav Smirnov, and Cédric Villani. Lindenstrauss received the Fields Medal for his work on measure rigidity in ergodic theory and its applications to number theory.
Researchers at the Mayo Clinic have developed an image-processing algorithm that can reduce the amount of radiation used in perfusion CT scanning, a technology used for diagnosing stroke, cancer, and possibly heart disease. The new adaptive algorithm compares multiple scans and effectively reduces image noise while preserving iodine si...
A new data mining algorithm called GOALIE reveals how biological processes are coordinated in time. The algorithm uses gene expression data to reconstruct temporal models of cellular processes, improving our understanding of cell division and metabolism.
Engineers at Boston University have developed mathematical algorithms that can make nearly optimal decisions under realistic model conditions. This approach automates functions such as task partitioning and re-planning, reducing human supervision needed for UAVs.
Researchers at NC State University developed a new analytical method to process large amounts of information efficiently. The algorithm allows for targeted collection of relevant data, minimizing errors and reducing data storage needs.
The MilkyWay@Home project uses volunteer computers from around the world to study dwarf galaxy movements and provide new details on dark matter. The project, which began in 2006, has surpassed one petaflop of computing speed and is the fastest computing project on the BOINC platform.
Researchers developed an innovative technique, Intelligent Outlier Detection Algorithm (IODA), to detect errors in sensitive technological systems. The algorithm can identify faulty readings or problems associated with failing sensors and isolate bad data.
Researchers developed an algorithm to target local scanning worms, which can evade worm defenses. The algorithm estimates susceptible host population size and quarantines the worm if scans exceed a threshold value.
Researchers at Berkeley Lab developed a molecular worm algorithm to automatically analyze structures, speeding up material screening. The algorithm provides a realistic depiction of molecule geometry, allowing for more accurate predictions of catalysis and chemical reactions.
Researchers at Kansas State University developed an algorithm that enables next-generation smart adaptive radios and makes radars more effective by customizing the coding used in transmitting signals. The algorithm can be applied to various systems employing spread spectrum techniques, including cellular, satellite, and wireless networks.
A new report by NIST demonstrates that iris recognition algorithms can maintain their accuracy and interoperability with compact images, affirming their potential for large-scale identity management applications. The study's results show that two compression standards, JPEG2000 and PNG format, performed well in preserving image quality.
Researchers in Japan developed a computer program that accurately predicts patients' risk of dying at the time of an emergency call. The algorithm achieved over 80% sensitivity and has potential to prioritize ambulances for critical cases.
Researchers at Yale University have observed and tracked E. coli bacteria moving in a liquid medium with a motion similar to a kayak paddle. The study, published in Physical Review Letters, provides new insights into bacterial movement and may lead to breakthroughs in preventing bacterial migration and sickness.
Researchers analyzed 86 patients with nasal fractures, finding no significant difference in revision rates or patient satisfaction between closed and open treatments when done correctly. A treatment algorithm based on fracture type and degree of septal deviation can help surgeons choose the best approach for each individual case.
The Stanford camera, also known as Frankencamera, is an open-source digital camera that will give programmers the chance to create software that teaches cameras new tricks. With the camera's operating software made available publicly, users can continuously improve it, adding their own algorithms to process raw images in innovative ways.
Researchers have developed a way to diagnose networking delays as short as tens of millionths of seconds, which can lead to significant financial losses. The Lossy Difference Aggregator approach offers the possibility of fine-grained delay and loss tracking with minimal cost and performance penalty.
Dr. Scott Craver and his team are investigating detectors resistant to adversary algorithms, as well as audio, image, and video fingerprinting techniques that embed imperceptible labels.
Scientists at the University of Georgia have created a two-step computer simulation using the Wang-Landau algorithm to study how glycophorin A folds into its functional shape. The research reveals that the process is driven by a subtle interplay between multiple types of interactions, providing insights into membrane protein folding.
Researchers have created a game called Magic Bullet to make labelling CAPTCHA images more efficient. Players can label images at a rate of up to 1,500 per hour, significantly improving the accuracy of human labels and helping to stay ahead of spammers.
Researchers developed a technique inspired by the human eye to improve object recognition, surveillance, and shape reconstruction. The new algorithm uses trust search regions to quickly identify objects with nearly double the accuracy of previous methods.
Researchers at Cornell University have developed a computer algorithm that can simulate the sounds of water and other fluids, allowing for more realistic graphics simulations. The algorithm is based on physics-based simulations and uses computational models to generate sound waves in response to vibrations and movements.
Huygens, the Dutch national supercomputer, establishes a new world record in Go by winning a 19x19 competition with a 7-stones handicap. The victory is made possible by the French software MoGo TITAN, which runs on Huygens and utilizes the Monte-Carlo Tree Search algorithm.
A new approach based on mathematical theory of elliptic curves has emerged as a leading candidate for more efficient cryptography. Elliptic curve cryptography provides the same level of security as 1024-bit keys for RSA with smaller key sizes, making it computationally more efficient.
A new computer algorithm assesses mammographic breast density by automatically segmenting breast tissue from background, removing muscle, and using region growing to determine dense areas. This method provides objective and accurate recommendations on further imaging or monitoring, reducing variability among radiologists.
Researchers at Caltech have successfully created a system using DNA origami seeds that can direct the self-assembled growth of DNA tiles into precise forms. This breakthrough demonstrates unprecedented control over information-directed molecular self-assembly, paving the way for future applications in technology and materials science.
A computer program using raw observational data teases out fundamental physical laws, particularly for biological systems that have eluded detection. The breakthrough uses a computational process inspired by biological evolution to analyze data sets and identify invariant patterns.
A study published in the International Journal of Biometrics proposes using knee X-rays for biometric validation, providing a unique and secure method for identifying individuals. The Wnd-charm algorithm rapidly analyzes X-ray images to identify specific knees and match them with a database, making it difficult for fraudsters to spoof ...
Researchers Johannes Soeding and Andreas Biegert have developed a new method called CS-BLAST that takes into account the sequence context to improve similarity searches. This approach can identify twice as many distant relatives of proteins compared to traditional BLAST, leading to better insights into gene and protein functions.
A Duke University-led team has developed a computer program that can redesign enzymes to produce natural antibiotics. The algorithm, called K*, sorts through possible shapes and changes of the key enzyme that produces gramicidin S, a natural antibiotic. This new technique may pave the way for more automated redesign of old drugs.
Scientists have developed a new equation of state for seawater based on Absolute Salinity, which is expected to provide more accurate thermodynamic calculations. The new approach aims to improve the understanding of seawater's properties and its role in the global heat engine.
A new mathematical model developed at Duke University could help robots navigate strange surroundings to find hidden explosives. The model, inspired by the popular board game CLUE, uses a strategy of selecting movements and optimizing information incorporation to achieve success.
A team of Duke University engineers has developed a new type of cloaking device using complex mathematical algorithms to guide the design and fabrication of exotic composite materials. The device successfully cloaks electromagnetic waves, bending them around an object to create an 'engineered mirage'.
Researchers at University of California, Davis have developed a 2-pronged approach to ward off self-propagating infectious programs known as worms. The strategy enables computers to share information about anomalous activity and use an algorithm to weigh the cost of infection against the benefit of staying online.
A new algorithm developed by scientists at University of California, Davis, can extract features and patterns from huge data sets using minimal computing power. The algorithm has been successfully applied to analyze complex phenomena represented by billions of data points.
Brent Heeringa's research focuses on three specific problems: optimal algorithms for hierarchical data, dynamic models of information organization, and new models for categorical data. His project aims to create systems where desired information is quick and easy to find.
A team of Berkeley Lab researchers won a prestigious Gordon Bell Prize for their algorithm innovation in high-performance computing. They developed the Linearly Scaling 3D Fragment (LS3DF) method to efficiently simulate the behavior of nanostructures, achieving impressive performance and scalability.
A team of Carnegie Mellon researchers has developed a new computational model that sheds light on the brain's processing of natural scenes. The model analyzes patterns in images and statistically characterizes them to determine which patterns are most likely associated with each other, enabling the brain to make sense of visual stimuli.