Articles tagged with Network Modeling
A recent study published in Radiology: Artificial Intelligence found significant racial and sex-related biases in an AI chest X-ray foundation model, affecting its performance across patient subgroups. The researchers highlighted the need for comprehensive bias analysis to ensure diversity and representativeness in dataset collection.
A new study found that living donor liver transplant access optimizes the timing of transplant for older, frail cirrhosis patients. Patients with moderate to severe frailty, short stature, and low MELD scores benefited from LDLT, reducing waitlist mortality and improving post-transplant outcomes.
The UMass Amherst-UT Austin team has been awarded $27.5 million by the CDC to establish the nation's first disease outbreak response and modeling network. The $250 million program will support decision makers during public health emergencies like COVID-19.
The CDC has awarded a coalition led by UC San Diego researchers $17.5 million to develop innovative tools and networks for rapid response to emerging disease outbreaks. The project aims to integrate forecasting and analytic approaches with data to inform public health decisions.
Researchers have successfully solved a problem in graph theory that has attracted attention from within the field. The team's research involves packing coloring, which deals with labelling parts of a graph to comply with certain rules and avoid specific conflicts.
Researchers found that adding niobium oxide to silicate glass increases bond density and connectivity, enhancing mechanical and thermal stability. This discovery could lead to the development of innovative glass formulations for various applications, including optics, medicine, and data transmission.
Wogrin aims to improve data aggregation and create more meaningful models with the same computing power, resulting in suboptimal investment decisions and costly restructuring of energy systems. Her research approach takes into account different supply situations, enabling compressed and differentiated model data.
Researchers used 3D simulations on Stampede2 to model the flow of HAT-P-32b's atmosphere, revealing a gigantic helium gas tail. The planet is losing significant atmospheric mass, which could help explain the mystery of intermediate-mass planets.
Researchers have developed a novel neural network approach to design brand new proteins with unique arrangements and dynamic functionalities. The method combines attention neural networks with graph neural networks to predict existing protein properties and envision new proteins that nature has not yet devised.
A hybrid system of electronic encoding and diffractive optical decoding transmits optical information with high fidelity through random, unknown diffusers. The system outperforms traditional approaches that only utilize a diffractive optical network or an electronic neural network for optical information transfer.
A new methodology, developed by USGS researchers, integrates earthquake-induced ground failures with ground shaking hazards to provide a comprehensive forecast for pipeline damage. This approach is crucial for identifying potential damage in buried gas pipelines covering hundreds of kilometers.
Researchers at EPFL have developed the Backtracking Dynamical Cavity Method (BDCM) to study disordered systems, which are found in materials science, climate, and social networks. By tracing steps backward from stable points, the BDCM provides valuable insights into complex system dynamics.
Researchers found a lognormal distribution of neuron densities in mammalian brains, influencing network connectivity and potentially promoting efficient information transmission. The discovery is relevant for modeling the brain accurately and designing brain-inspired technology.
A team led by Dr. Zixiang Xiong at Texas A&M University aims to understand the fundamental limits of learned source coding, a machine learning-based data compression method. They hope to develop more powerful compression methods for efficient use of wireless communication and less energy consumption by mobile devices.
Researchers propose a hypothesis that astrocytes, non-neuronal cells in the brain, can perform core computation as transformers, providing insights into human brain function and machine learning success. This discovery could spark future neuroscience research and help explain transformer performance across complex tasks.
A study published in PNAS found that government-mandated external moderation is likely to be effective in limiting harm on fast-paced platforms like X. The researchers examined the dynamics of content dissemination and found that a lower half-life means most harm happens right after content is posted.
A new study by Paula Mayer proposes a practical tool for identifying areas of high human-bear conflict in the Abruzzo region. The model considers factors such as habitat suitability, migration corridors, and human-made food resources to inform local measures promoting coexistence.
Researchers discovered intricate neural connections in Octopus vulgaris, a model organism for studying memory acquisition networks. The findings challenge traditional models of neural network functionality, revealing an evolutionary adaptation that underlies the octopus's unique cognitive prowess.
Researchers used a mathematical theory called the free energy principle to predict how real neural networks learn and organize themselves. The study successfully mimicked this process in rat embryo neurons grown in a culture dish, demonstrating the principle's guiding force behind biological neural network learning.
A new complex-domain neural network enhances large-scale coherent imaging by exploiting latent coupling information between amplitude and phase components. The technique reduces exposure time and data volume significantly while maintaining high-quality reconstructions.
A cross-sectional survey of 121 patients found that 90% were very satisfied or satisfied with family physician care, and 89% with allied health team care. Comparing to previous experiences, 75% believed IVC encounters were better or same as in-person encounters.
GOBI overcomes model-free and model-based inference method limitations by introducing an easily testable condition for a general monotonic ODE model to reproduce time-series data. It successfully infers positive and negative regulations in various networks, distinguishing between direct and indirect causation.
A new geometric deep learning model called GFCN has been developed to detect stroke lesions in brain imaging scans. The model leverages rich geometric information to segment brain tissue and achieves higher segmentation performance than other neural network architectures.
The study proposes a novel network slicing planning and handover technique applicable to next-generation low-earth orbit (LEO) satellite networks. The developed technique can provide flexible services according to user needs, enabling the adoption of 6G-era demands such as VR/AR and autonomous driving.
A team of researchers from Kyoto University and international institutions has developed a mathematical solution to the temporal asymmetry of nonequilibrium disordered Ising networks. This breakthrough offers insights into the behavior of biological systems, machine learning, and AI tools.
A team of scientists identified the dorsal medial prefrontal cortex as a key region in predicting rumination, which is linked to depression. The study's findings suggest that dynamic connectivity between brain regions can be used to decode rumination patterns.
Researchers at MD Anderson Cancer Center have engineered a new model of aggressive renal cell carcinoma, highlighting molecular targets and genomic events that trigger chromosomal instability. The loss of interferon receptor genes plays a pivotal role in allowing cancer cells to become tolerant of chromosomal instability.
BioAutoMATED is an all-in-one AutoML platform designed for biologists, enabling easy analysis and interpretation of biological sequences. The platform uses three existing AutoML tools to generate models that can predict biological functions from sequence information.
Researchers from Complexity Science Hub highlight similarities in models used by economists and physicists to analyze financial markets. The study aims to create an overview of these models, promoting interdisciplinary collaboration and avoiding research that gets lost in translation.
The study identified two relevant circuits in patients with disorders of consciousness (DoC), including the thalamo-frontotemporal region and the posterior cortical region. These findings bring a new understanding of brain networks and could improve diagnosis and treatment for people suffering from DoC.
A recent study by the Complexity Science Hub reveals interdependencies in the global food supply chain, uncovering profound indirect effects of the Russia-Ukraine conflict on product availability worldwide. The study found that indirect effects often exceeded direct effects, with losses up to 85% in maize and 89% in edible oils.
A community-wide program in Washington, D.C., supports adolescent mothers, caregivers, and community members, addressing barriers to childcare and education. The 'collective impact' model has improved access to essential programs, including housing and food security, for over 550 young parents.
Researchers developed a new spectropolarimetric imaging technique called DIP-SP, which integrates a passive polarization modulator into an imaging spectrometer. This approach enables high-dimensional information capture from incomplete measurements and significantly improves image quality.
A study by Dartmouth's Geisel School of Medicine found that Medicare fraud in home healthcare billing spread rapidly across U.S. regions between 2002 and 2009, driven by characteristics such as shared patients, high expenditures, and rapid growth in the number of home health agencies. The researchers developed a novel network analysis ...
EPFL researchers used Chat-GPT to design a working robotic tomato harvester, showcasing the AI tool's potential for collaborating with humans in robotic design. The study highlights opportunities and risks of applying artificial intelligence to robotics, emphasizing the need for careful evaluation of LLMs' role in design.
Researchers built personalized brain network models to simulate brain dynamics and found that people with higher fluid intelligence took more time to solve difficult tasks compared to those with lower FI. This suggests that a synchronized brain is better at solving problems but not necessarily faster.
Researchers developed an AI system, Geneformer, to predict how disruptions in human gene connections cause disease. The model, trained on data from thousands of genes, can identify potential drug targets for diseases like heart disease and cancer.
Researchers propose a deep neural network-based method for calibrating 4-quadrant analog solar sensors, reducing errors by up to 0.25° (3σ). The approach uses cubic surface fitting and deep feedforward neural networks to approximate the actual error model and correct errors effectively.
A recent study on canine brain networks has provided insights into the evolution of human brain function, revealing that the cingulate cortex played a central role in mammalian brain development. The research used fMRI to analyze brain activity in dogs and identified functional networks that differ from those in humans.
A new study by the Complexity Science Hub recommends aligning quotas with an inclusive culture to improve women's representation in science. The research finds that even extreme quotas are not sufficient to increase minorities' visibility in top ranks, highlighting the need for behavioral interventions and regulation to overcome biases.
A machine learning model helps explain how brains recognize the meaning of communication sounds, such as animal calls or spoken words. The study models sound-processing networks in social animals' brains and demonstrates their ability to distinguish between different sound categories.
Researchers developed GAME-Net, a graph neural network that rapidly evaluates adsorption energy for large molecules like plastics and biomass. The model achieves accuracy comparable to density functional theory (DFT) while utilizing simple molecular representations.
Researchers chart course for generalist medical AI, capable of performing complex tasks in a wide range of scenarios. The model can integrate multiple data types and apply existing knowledge to new contexts, offering comprehensive solutions to alleviate clinician burnout and reduce clinical errors.
Computational heuristics like Grey Wolf Optimizer and Whale Optimization Algorithm offer fast solutions to complex supply chain problems, providing accurate results within a shorter time frame. The research suggests these tools can be a valuable addition to supply chain management, especially in responding to unexpected disruptions.
Researchers study how choice of network framework affects synchronization, finding that hypergraphs promote easier synchronization than simplicial complexes. The choice of representation can influence the outcome, suggesting structural heterogeneity is key to understanding collective dynamics.
A recent study found that the Epic sepsis model's accuracy in predicting sepsis onset depends on hospital factors such as sepsis incidence and multiple health conditions. The model performed worse in hospitals with higher rates of these conditions, suggesting it may be more useful in lower-acuity settings.
A study published in Circulation Research identifies the FHL5 gene as a key regulator of vascular disease, including heart attacks and aneurysms. The discovery advances our understanding of the underlying causes of vascular disease and provides new insights into genetic risk factors.
A new theorem helps system designers make informed decisions about consistency and availability trade-offs in networks. The CAL theorem provides a quantitative relationship between these factors, allowing for more efficient and reliable network designs.
OncoMerge uses genetic data to analyze tumor activity and predict future changes. The software detects abnormal gene fusions and mutations affecting protein expression and gene copy numbers, improving the accuracy of cancer modeling predictions.
Researchers developed a model to track COVID-19 data, predicting transmission and informing health surveillance systems. The model successfully predicted the spread of COVID-19 in Cali, Colombia, highlighting the importance of high-resolution data in understanding virus dynamics.
Scientists developed a modeling technique to study urban traffic flows and verified it with real-world data from Shanghai. They discovered that Zhonghuan Road is a potential bottleneck that could lead to cascading failure of the entire urban traffic system.
A new approach to describing network connections can help predict system strong and weak points, crucial for understanding disease spread and communication networks. Researchers found that mapping hierarchies and incoherence within a system enables prediction of strong and weak connections.
Researchers developed AI models based on UNet and MobileNet architectures to analyze standardized abnormalities in CT images, accurately identifying object presence and confidence. These models achieved an absolute percentage error of less than 5 percent, comparable to human professionals.
A novel AI architecture, relational reasoning network, accurately identifies anatomical landmarks in CT scans for orthodontic treatments. The model learns spatial relationships between landmarks without explicit image segmentation, achieving accuracy comparable to conventional methods.
Researchers at Stanford University have developed a novel AI-powered approach to analyzing traumatic brain injury, using artificial intelligence to identify the most accurate model of mechanical stress on the brain. This breakthrough could lead to better understanding of when concussions lead to lasting brain damage and inspire new pro...
A novel deep learning-based forecasting model predicts uncertain parameters related to renewable energy sources, their energy demand, and market prices. The model demonstrates improved prediction accuracy and efficiency compared to existing methods.
Researchers seek to develop algorithms providing meaningful explanations for AI decision-making, enabling higher human trust and adoption in fields like science. The project focuses on symbolic reasoning and estimating explanation accuracy, addressing the need for transparent AI systems.
A new technique maps the effects of fire-induced permafrost thaw in Alaska, revealing widespread topographic change and vegetation shifts. The study used a machine learning-based approach to quantify thaw settlement across 3 million acres of land, with results showing a significant loss of evergreen forest and shrubland encroachment.
Researchers at MIT developed a technique to improve machine-learning models' reliability without requiring additional data or extensive computing resources. The method uses a simpler companion model to estimate uncertainty, enabling more effective uncertainty quantification.
A team of researchers developed a model-free approach using deep reinforcement learning to optimize estimation of multiple parameters in quantum sensors. The protocol achieved significantly better estimations compared to nonadaptive strategies, demonstrating enhanced performance in resource-limited regimes.