Articles tagged with Network Modeling
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.
Researchers utilized the Chemistry42 platform to generate novel molecular structures and identified a hit molecule for CDK20, a promising target for hepatocellular carcinoma. The platform's customizable reward function and generative models enabled efficient design and optimization of molecules.
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.
A systematic review found that higher vitamin D intake was associated with a 15% decreased likelihood of developing type 2 diabetes in adults with prediabetes. Inexpensive vitamin D supplementation could delay the development of diabetes in over 10 million people worldwide.
Researchers developed a neural network model that uses terahertz time-domain spectroscopy data to predict burn healing outcomes with high accuracy. The new approach improves upon existing methods by reducing training data requirements, making it more practical for processing large clinical trials.
A new study finds that autonomous vehicles could consume enough energy to generate significant greenhouse gas emissions, highlighting the need for rapid advancements in hardware efficiency. To mitigate this, researchers recommend more efficient autonomous vehicles with smaller carbon footprints.
A recent study uses sparsification to identify critical links in a network, reducing computation time for simulating disease spread. The technique preserves overall dynamics while removing non-essential edges, enabling faster simulations of large-scale pandemics.
Researchers found no evidence of a critical mass needed to start and maintain new research fields. Instead, pioneering regions with early investment can establish dominance. However, late-comers face significant costs to catch up, as seen in China's semiconductor science, where strategic interventions over decades led to a dominant role.
The article discusses the use of Rapamycin in the context of Pascal's Wager, a philosophical framework used to justify beliefs. ChatGPT's generative pre-trained transformer model provides an exhaustive research perspective on the pros and cons of taking Rapamycin for anti-aging purposes.
A new study finds that global warming overshoots could trigger climate tipping events, even if temperatures are limited to 1.5 degrees Celsius. The risk of triggering one or more tipping points would still be over 50 percent in such scenarios.
Researchers at Nara Institute of Science and Technology have developed a method to measure the congruence between contributor networks and library dependencies in open-source software ecosystems. By analyzing over 5.3 million change commits across 107,242 libraries, they found that libraries with high levels of matching contributions a...
A team of researchers proposes an intelligent routing scheme to optimize link load balancing in software-defined networks, achieving significant performance improvements. Their algorithms outperform traditional methods, reducing maximum bandwidth by 24.6 percent in real-world topologies.
A new deep learning model using conditional generative adversarial networks (cGANs) has been developed to diagnose bowel ischemia. The model achieved an accuracy of over 93% in segmenting ischemic intestine images, surpassing current subjective methods.
Researchers trained a mouse brain model to solve visual tasks, achieving superior robustness and performance compared to traditional neural networks. The model's unique coding properties enable it to cope with errors and unexpected input, making it a promising tool for advances in neuromorphic computing.
Deep learning models can become less accurate in recognizing specific categories of images, sounds, or text after network pruning. Researchers demonstrate a technique to address this challenge, improving the fairness of deep learning models.
Researchers at MIT have developed a machine-learning model that captures how sounds propagate through spaces, allowing for accurate visual renderings of rooms. This technique has potential applications in virtual and augmented reality, as well as improving AI agents' understanding of their environment.
A team of CSU researchers has designed a model that can predict which buildings will survive a wildfire, allowing for more effective fire mitigation strategies. By analyzing community networks and incorporating graph theory, the model achieves accuracy rates of up to 86% in predicting building survival.
Researchers developed a new method to represent cell communication using graph neural networks, which uncovers the effects of tissue niche composition on gene expression. The node-centric expression models (NCEMs) identify cell-cell dependencies and molecular processes underlying cell communication.
Researchers at MIT have developed a new method that uses optics to accelerate machine-learning computations on low-power devices. By encoding model components onto light waves, data can be transmitted rapidly and computations performed quickly, leading to over a hundredfold improvement in energy efficiency.
A new study by Helena Miton and Simon DeDeo presents a model for the transmission of tacit knowledge, which is passed down with limited specification. The model captures how learners overcome constraints to succeed in complex practices, predicting stability over time with minimal information.
A new CNN framework, PE-Net, is proposed for predicting machine remaining useful life (RUL) accurately. The framework uses a novel architecture with small-sized one-dimensional convolution kernels and deep networks to learn features from input time series signals.
A lightweight 'non-local' generic face reconstruction model using the Speckle-Transformer (SpT) Unet is implemented for highly accurate and energy-efficient processing of speckle reconstructions. The network achieves strong comparative performance with existing methods, exceeding 0.989 in Pearson correlation coefficient.
A team of Illinois Tech researchers used machine learning to estimate the age and gender of individual users with high accuracy, raising questions about data security and privacy. The study highlights the need for better regulations and best practices to protect personal information from being misused.
Researchers developed computational models to identify massage businesses at risk of violating laws related to sex and labor trafficking. The models provide probability scores on the likelihood that a business is engaged in illegal activity, allowing law enforcement and organizations to prioritize investigations.
Researchers analyzed LNS's tail latency and low entropy benefits compared to mTCP and Linux network stacks. The study revealed that fulldatapath prioritized processing and full-path zero-copy are primary factors for high performance, improving tail latency by up to 5.5 times.
A York University study found that deep convolutional neural networks (DCNNs) fail to capture the configural nature of human shape perception, which could be dangerous in traffic video safety systems. The researchers discovered that while humans use configural shape perception, DCNNs take 'shortcuts' and are insensitive to this aspect.
A team of scientists developed a chip that simulates the human lung's breathing pattern, allowing them to visualize and analyze the flow of air and particulates through the alveoli. They found distinct flow patterns for different generations of the bronchial network, shedding light on respiratory diseases such as emphysema and COPD.
Researchers from Singapore University of Technology and Design (SUTD) have developed a new Brain-Inspired Replay model that enables continual learning in edge computing systems without storing data. This approach achieves state-of-the-art accuracy and high energy efficiency, overcoming the stability-plasticity issue in traditional models.
A new method for generating realistic images in driving simulations uses machine learning to improve visual fidelity. This enables better testing of driverless cars and study of driver distraction, ultimately enhancing safety and interaction between humans and AI on the road.
Scientists have developed network-based models to prioritize novel viruses for their zoonotic transmission risk, with coronaviruses found to be the riskiest. The models also ranked paramyxoviruses, such as those causing measles and respiratory tract infections, as high priorities.
Researchers at MIT developed an AI model that can detect Parkinson's disease from breathing patterns, using a neural network to assess the presence and severity of the condition. The device is non-invasive and can be used in patients' homes without any bodily contact.
A new optimization model aims to improve the efficiency of bike-sharing systems by predicting user demand and adjusting service operations accordingly. The model has shown promising results, reducing problems by 41% compared to no rebalancing.
Researchers at NC State University have developed a cooperative distributed algorithm that allows autonomous vehicle software to make calculations more quickly, enabling real-time navigation of complex merging scenarios. The approach improves both traffic flow and safety, with zero incidents in simulations.
A new predictive network model estimates how much dissonance people experience when holding conflicting beliefs about a topic. This approach can help determine who will change their minds about contentious scientific issues when presented with evidence-based information.
A study analyzing over 2 billion Reddit comments found that 16.11% of users publish toxic posts and 13.28% of users publish toxic comments, with a positive correlation between community growth and increased toxicity
A research team used game theory to analyze cooperation in networks and found that networks with a high level of cooperation can emerge if individuals take a clear-cut position against free riders. The study also showed that if contributors leave an environment too quickly, it leads to a lower level of cooperation.
A team at the Complexity Science Hub Vienna mapped an entire nation's supply chain network using mobile phone data, predicting systemic risk and resilience. The model can be easily implemented by other countries and provides a detailed view of national economic behavior on a daily timescale.
A new study reveals that droughts in the Amazon rainforest can lead to a ripple effect, causing widespread damage and biodiversity loss. The research team used network analysis to understand the complex workings of the forest's moisture recycling system.
A team of scientists developed a computational model that explains Italy's town distribution using only a small set of mathematical equations and a map of the landscape. The model simulates how population and road networks interact, demonstrating that landscape alone is insufficient to explain population distribution.
A team of researchers from Boston University and EcoHealth Alliance will develop models to predict disease emergence and spread. They aim to identify location hotspots for pathogen emergence and determine the most effective pandemic mitigation strategies using data from COVID-19, H1N1 flu, and Ebola Virus Disease.
A research team from the Complexity Science Hub Vienna has developed a stress test to identify weaknesses and strengths in healthcare systems. They used data from Austria to show how many resident physicians can drop out before patients don't find a new doctor within reasonable distance, highlighting regional differences in resilience.
Researchers at OU's CQRT are developing quantum synchronization and organization using multiple experimental approaches. They aim to create a quantum network and better understand collective interactions, with potential implications for network synchronization and electrical power systems.
The partnership aims to enhance early detection and reporting for emerging diseases, building a robust decentralized global surveillance network. The collaboration will focus on advancing equitable data-sharing practices, bolstering epidemiological and genomic surveillance in low- and middle-income countries.
Using the Stampede2 supercomputer, researchers have developed a deep learning model that predicts the properties of over 370,000 high-entropy alloy compositions. The study also applied association rule mining to discover design rules for high-entropy alloy development and proposed several compositions for experimentalists to synthesize.
Researchers developed a novel computational method to control large complex networks using local information, addressing the challenge of controlling complex systems. The method considers computational time and information communication costs to produce optimal choices.
A team of researchers developed a model system to study individual differences in metabolism using C. elegans worms. They identified a novel metabolic condition linked to variation in the hphd-1 gene, which has implications for personalized medicine and tailoring dietary advice and disease treatment to an individual's genome sequence.
Researchers developed a novel convolutional neural network for facial expression recognition, outperforming conventional models while being computationally less expensive. The new model achieved an accuracy of 72.4% using only 58,000 parameters.
Researchers developed a convolutional neural network to identify fetuses with Down Syndrome from ultrasound images. The model achieved high accuracy, improving detection by over 15% compared to existing methods. Non-invasive screening could become a convenient and inexpensive tool for early pregnancy diagnosis.
An international research team applied a new dynamical model and exceptional historical data to examine major explanations for cultural complexity evolution. The study found little support for many influential theories, including the transition to agriculture and conflict theories.
A new AI system uses artificial neural networks to recognize objects more accurately and stably, despite changing visual inputs. The system mimics human eye movements to improve machine vision capabilities, reducing errors in self-driving cars and other applications.
Researchers used microcircuit models of basal ganglia and thalamus areas to create multiscale models of Parkinson's patient and healthy control brain. They found that in-silico deep brain stimulation could normalize decreased firing rates in subcortical regions, but also caused differential activity in the motor cortex.
Researchers at Portland State University created a two-stage stochastic programming model to optimize e-scooter placement, charging, and rebalancing in Tucson. The model uses data from the City of Tucson to answer questions on demand uncertainty, idle periods, and customer satisfaction.
Researchers developed a model inspired by vascular biophysics to examine the stability of coastal channel loops. They found that strong interplay between rivers and tides is essential for loop stability, and that flows must constantly rearrange themselves to keep them open.
Researchers developed a computational model to determine optimal places for electric vehicle charging facilities and powerful stations without straining the local power grid. The model considers travel flow, user demand, and regional power infrastructure needs.
A new neurobiological model provides a unified framework for understanding various forms of creativity, including abstract thinking, improvisation, and divergent problem-solving. The model suggests that different brain areas are activated depending on the type of creativity, with dopamine playing a crucial role in controlling and optim...
A team from Nagoya University created an artificial neural network model that performed the delayed matching-to-sample task and analyzed its behavior. The model was able to evolve to exhibit human-like metamemory, adapting to its environment by learning and evolving. This breakthrough aims to create machines with memories like humans.
Research suggests that during deep sleep, neurons representing related items fire in close temporal order, triggering synaptic plasticity and forming strong connections. This process strengthens or creates new relational memories, which can be essential for learning connections between objects or people.
A recent study out of the Complexity Science Hub Vienna developed a mathematical and computational framework for analysing neural activity in C. elegans, a tiny worm used to study neural activity. The study proposes a way to unmask the roles of neurons by using more natural perturbations.