Articles tagged with Algorithms
A new mathematical method provides a better way to analyze noise by visualizing areas with no sound. Researchers credit this approach for its ability to tease out specific sounds from complex white noise.
Dr. Éva Tardos received the George B. Dantzig Prize for her groundbreaking work on network-flow algorithms, approximation algorithms, and combinatorial auctions. Her research focuses on efficient methods for solving optimization problems in graphs and networks.
The recipients of the 2006 Lagrange Prize are recognized for their groundbreaking papers on nonlinear programming without a penalty function and global convergence of a filter-SQP algorithm. These works introduced the filter idea, which has led to the development of effective nonlinear optimization codes.
A Florida State University researcher is creating unmanned ground vehicles to patrol areas without putting soldiers in harm's way. The Center for Intelligent Systems, Control, and Robotics is also working on algorithms that enable autonomous devices, such as wheelchairs and self-parking cars.
Researchers used a technique to analyze political book sales and blog networks, finding distinct communities and cohesion within groups. The algorithm identified left- and right-wing book groups and blogs, with rare connections between them, suggesting polarization and factional strength.
Indiana University researchers have created an Inauthentic Paper Detector (IPD) that uses compression algorithms to identify and distinguish between meaningful and meaningless technical texts. The system was inspired by a 2004 prank where MIT students submitted fake research papers, which were accepted without review.
Researchers develop two mathematical tools to analyze and optimize tensegrity systems in various applications. The optimization algorithm helps maximize strength while minimizing weight, leading to potential breakthroughs in bridge, building, and aircraft design.
Computer science has a profound impact on various scientific fields, including biology and physics. Chazelle emphasizes the importance of understanding algorithms in capturing complex realities, rather than traditional mathematical formulas.
A new algorithm called Automatic Distillation of Structure (ADIOS) has been developed to learn complex patterns in raw texts, including transcripts of parents' speech directed at young children. The algorithm can generate grammatical new sentences and prove useful in fields like bioinformatics.
Researchers from the California Institute, led by Professor Henrik Wann Jensen, unveiled innovative computer graphics techniques at SIGGRAPH. Their work includes a new method for real-time relighting of scenes illuminated by local light sources and an efficient technique to account for diffusion in multi-layered translucent materials.
The Common Message Authentication Code (CMAC) is a new authentication mechanism that uses block ciphers to verify the integrity of digital data. Developed by Japanese scientists Tetsu Iwata and Kaoru Kurosawa, CMAC is part of NIST's ongoing effort to update block cipher-based algorithms.
Artificial tactile sensors are being developed to mimic human fingers, providing detailed information on object shape and texture. The new technology could enable robots to handle delicate objects with increased precision and force control.
A study by Purdue University researchers found that pi's randomness is acceptable for data encryption and physics problems, but not always the best. Some commercially available random number generators outperformed pi in certain tests, suggesting that the quality of the algorithm used can impact randomness.
A new algorithm could significantly improve transplant matches for patients with end-stage kidney disease, with a national program resulting in 47.7% more transplants and better graft survival rates compared to regional programs. The study also found that highly sensitized patients would benefit six-fold from the optimized algorithm.
A national matching program for kidney donors and recipients can significantly increase the number of transplants, improve match quality and survival rates, while reducing travel costs.
Researchers from the University of Toronto developed a new method for predicting mammalian gene function by analyzing RNA expression patterns. They used a support vector machine algorithm to analyze over 40,000 known mouse mRNAs and predicted functions for thousands of genes, with high accuracy.
A team of researchers has discovered that microRNA miR-375 regulates insulin secretion, opening up new avenues for understanding and treating diabetes. The study's findings define a biological function for a mammalian microRNA gene and highlight the importance of collaboration between computation and experiment in modern biology.
Researchers have developed a mathematical algorithm called TANGO that can predict the likelihood of proteins sticking together incorrectly. This could lead to new diagnostic techniques for diseases caused by misfolded proteins, such as Alzheimer's and Parkinson's, and more efficient production of therapeutic proteins.
A Stanford researcher has devised a method to identify potential drug compounds using a network of over 150,000 home computers and innovative algorithms. The method accurately predicts how well molecules will bind to a given protein, which is crucial in drug development.
The three-year funding supports a joint effort to apply Multiple Dirichlet Series to analytic number theory, a central area of modern mathematics. This could lead to new insights into long-standing problems in number theory.
A study found that enzymes in plant cells can produce different products based on their location within the cell. The research, conducted by Brookhaven National Laboratory scientists, suggests that modifying an address signal on these enzymes could change their product output.
Researchers found that depression is often associated with an underlying autoimmune process affecting the thyroid gland. The study used thyroid function indicators to predict patients' responsiveness to antidepressants, achieving an almost 90% success rate.
Researchers developed an algorithm, Q5, using PCA and LDA to differentiate between healthy and diseased blood samples. The algorithm achieved virtually 100% accuracy for ovarian cancer and approximately 95% accuracy for prostate cancer.
Researchers at the University of Illinois at Urbana-Champaign have developed a geometric cluster algorithm that accelerates simulations of complex fluids. The new method can efficiently capture the motions of particles of different sizes, resolving a long-standing challenge in fluid simulation.
The ADOPT-A trial demonstrates the effectiveness of the AF Suppression algorithm in reducing symptomatic atrial fibrillation (AF) burden. The study found a significant 25% reduction in symptomatic AF episodes in patients using the algorithm compared to those with standard pacemaker indications.
Engineers at Ohio State University have developed a new software algorithm that models translucent objects and fluids to create soft, realistic shadows. The algorithm, called splatting, uses volume elements and 2D footprints to generate accurate shadows with minimal data storage requirements.
The University of California, San Diego (UCSD) has received a $600,000 grant from the Technical Support Working Group (TSWG) to develop advanced video surveillance systems for public safety. The project aims to improve the ability to detect and track potential terrorists using 'smart' cameras that can adapt to various lighting conditions.
A Stanford team has developed three techniques to speed up computing Google-style Web rankings, with speeds of up to five times faster than current methods. The techniques, including extrapolation, BlockRank, and Adaptive PageRank, can reduce computation time by 50-300% in realistic conditions.
A new method, KL clustering, has been identified as more effective than hierarchical clustering in separating gene expression data into dense clusters. The approach uses a novel measure of similarity to create well-separated clusters, offering insights into gene-gene interactions and biological processes.
Arjun Heimsath and Hany Farid developed a new method to create 3D models of remote regions using 2D photographs, making it easier for researchers to predict landslides and erosion rates. The technique has limitations, such as requiring clear images of the ground surface and a good point of view.
The new software tool analyzes engine models and extracts information on mechanically sound designs, enabling faster prediction of jet engine performance. By using an optimization algorithm, the tool calculates worst-case vibration levels, allowing engineers to determine the reliability of their designs quickly.
Scientists from the University of Minnesota and Duke University developed computer techniques that can automate image inpainting, restoring lost regions with neighboring available information. This technology has applications in surveillance images, enabling faster recovery of lost information and potential super resolution.
Researchers develop algorithm to straighten arcs in a plane without parts bumping into each other. The breakthrough applies to robotics, antennas on satellites, and protein chains, offering insights for engineers and biologists.
Researchers have designed an 'Energy Landscape Paving' method that circumvents the problems of the annealing algorithm, enabling fast and automatic solution-finding. This breakthrough could lead to better understanding of proteins' 3D nature and their functions, with potential applications in pharmaceuticals and materials development.
A recent study suggests that math constants like pi and log(2) could be normal, meaning their digits are random in a statistical sense. The research, led by David H. Bailey, finds that certain sequences in chaotic dynamics may underlie the normality of these constants.
Weixiong Zhang developed an algorithm to tackle the Traveling Salesman Problem (TSP), a classic planning and scheduling problem, and found it to be the best solution for half of the tested problems. The algorithm has real-world applications in logistics, business, and even biological data analysis.
Eight scientists have been selected for the Gerhard Hess prize for their groundbreaking work in humanities, biology, and medicine. The winners include Dr. Martin Wallraff for his critical edition of a Christian author's chronography, and Dr. Thomas Behr for developing radioimmunotherapies to treat tumours.
The PSLQ algorithm has found surprising relations in mathematics and physics, including a formula for calculating any binary digit of pi without prior calculations. This breakthrough uses high precision arithmetic to discover new mathematical relations among numbers.
Rescuers can locate avalanche victims four times faster using a robotic swarm search technique. Researchers at Sandia National Laboratories developed a computer program that enables humans to find buried skiers using mini-robots and simple radio equipment.
Researchers at the University of Delaware have developed a mathematical algorithm that can precisely control blood-sugar levels in people with Type I diabetes using an implantable insulin pump. The chip-based system predicts insulin needs based on glucose sensor data, reducing peak glucose levels by 44% and overshoot by 80%.
Jon Kleinberg, a computer scientist at Cornell University, is working on making the Internet more efficient by studying how information moves through networks. He aims to develop better routing algorithms that can handle congestion and optimize network capacity.
A study by researchers at UCSF Medical Center found that the electrocardiogram is highly inaccurate for patients with left bundle-branch block (LBBB), leading to under-treatment. This affects approximately 100,000 Americans with LBBB who experience myocardial infarction every year.
A team of engineers has developed a theory and algorithm that enables robots to adapt to changing conditions, solving a two-decade-old problem in intelligent control. The Max-Plus Algebra Model integrates low-level data sensing with high-level planning and decision-making processes, allowing robots to halt activity when necessary and r...
The Rice-Rutgers team developed a computer program that determines the optimal route for a salesman visiting 13,509 cities, exceeding previous limits of 2,892 cities. The program relies on sophisticated algorithms and parallel processing, with the code running to over a thousand pages if printed.
A new method for clustering data on computers has been developed by Prof. Eytan Domany, enabling the analysis of vast amounts of information without prior knowledge of its structure or categories. The algorithm mimics human intuition and can automatically identify clusters in various types of data.