Articles tagged with Data Points
A global analysis of over 2,300 seawater samples reveals human-made chemicals make up a significant portion of organic matter in coastal oceans. Industrial chemicals, including plastics and consumer products, dominate the anthropogenic chemical signal, persisting even 20 kilometers offshore.
Researchers from ISTA developed an algorithm that can extract and analyze information from the world’s most extensive biobank with unprecedented accuracy and speed. The method, dubbed gVAMP, enhances the framework's ability to extract complex information from the dataset at hand, providing a detailed overview of the effects on a trait ...
A new theory asserts that associative learning depends less on repetition and more on time passing between rewards, explaining why students who cram don't perform well. The study's findings could shift the way we look at learning and addiction, potentially speeding up artificial intelligence by borrowing from this new approach.
MIT researchers have developed a new approach to assess predictions with a spatial dimension, leading to more accurate forecasts in spatial prediction tasks. The new method provides better validations than classical methods, especially in cases where assumptions about data independence break down.
A new study by UC San Francisco found a marked increase in advanced prostate cancer rates across California since doctors stopped routine screening, with death rates plateauing in most regions. The study emphasizes the need for screening that can identify fatal tumors without raising false alarms.
A new technique identifies and removes specific points in a training dataset that contribute most to a model's failures on minority subgroups. This approach maintains the overall accuracy of the model while improving its performance regarding underrepresented groups.
A new study from Tel Aviv University uses smartphone data to predict wildfire risk, overcoming individual device errors by averaging large amounts of public data. The method provides valuable insights into wildfire evaluation, especially in remote areas lacking traditional weather stations.
Researchers have introduced a new AI calibration method called Thermometer, which enables efficient calibration of large language models for various tasks. This technique leverages temperature scaling to adjust a model's confidence and can generalize to new tasks without requiring additional labeled data.
Researchers used movement sensors to track thoroughbreds' stride patterns and identified miniscule changes associated with increased injury risk. The technology has been shown to prevent catastrophic injuries by alerting trainers and veterinarians to intervene before a major injury occurs, resulting in longer careers for affected horses.
A new Bayesian inference framework reduces data size by over 80% while achieving accurate modeling and control of optical power evolution. The approach enables simultaneous exploitation and exploration of a data space to identify suitable candidates for autonomous driving optical networks.
Researchers developed an algorithm that uses stochastic differential equations to identify key data points in noisy data, improving early warning signal accuracy. The approach can be applied to various domains, including ecology and mental health, and has been validated through simulations.
Researchers developed a 360-degree head-up display system that can project holographic representations of road obstacles, improving road safety. The system uses LiDAR data to create accurate and dynamic information about potential hazards.
Researchers found that green spaces alleviate extreme heat's negative impacts on human health, while densely packed buildings increase mortality risk. Urban design strategies incorporating different types of greenery are recommended to mitigate heatwave-associated mortality.
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.
A University of Copenhagen study examines how data is used in asylum decisions, revealing inconsistencies and discrepancies that can lead to applicants being deemed untrustworthy. The research highlights the opaque nature of the decision-making process and the need for critical evaluation of new data types.
A new study suggests that the format in which graphs are presented may be biasing people into being too optimistic or pessimistic about the trends the graphs display. Researchers found that participants made lower judgements when presented with bar charts instead of line graphs, even when the data was left unshaded.
A team led by Jianxi Gao developed equations to measure and compare distances to tipping points across various mutualistic systems, allowing for the triage of declining ecosystems. This method enables the identification of areas that need intervention most urgently, providing a critical sense of urgency to mobilize resources.
A new study uses high-resolution regional model experiments to explore how lake surface temperatures may affect the climate of the Great Lakes region. Small differences in lake surface temperatures can have a significant impact on summer climate and fuel extreme weather events.
A recent study combined elevation data of the Antarctic ice sheet from two different satellites, CryoSat-2 and Sentinel-3, to obtain a more accurate map of latest elevation changes. The results showed that the average elevation of the ice sheets had decreased at a rate of 4.3 cm/year during 2016-2019.
Researchers developed an EHR-embedded card study to replace traditional paper-based methods in clinic-based research. The digital tool increased efficiency and reduced labor costs, as each participant's responses were linked to clinic and patient data.
A new study found that the fourth COVID-19 vaccine reduces the risk of death among elderly individuals in long-term care facilities by 72%. The study, which included approximately 40,000 residents, also showed a significant reduction in hospitalizations and infections.
Researchers developed a lidar-based system for smart cars to recognize objects more accurately than cameras. The system uses a grid map to divide the field of view into regions containing individual objects.
Researchers developed a system to estimate the probability of a majority response being wrong, eliminating unnecessary task assignments and increasing efficiency in crowdsourcing projects. This method can be applied to various binary classification tasks, reducing labor and improving accuracy.
Alternative credit scoring systems present challenges not covered by existing regulations, leaving many Americans unable to access the credit system. These systems can be opaque and ripe for abusive lending practices, making it difficult for consumers to contest erroneous information.
Researchers propose an open platform to integrate data collection, processing, and publication, ensuring FAIR (findable, accessible, interoperable, reusable) data and enabling machine-learning algorithms to learn from failed experiments. The platform aims to bridge the gap between data and machine learning in chemistry.
Researchers at MIT created a process called DualFair that can remove bias from data used to train machine-learning models. The method tackles both label bias and selection bias, significantly reducing discrimination in loan predictions while maintaining high accuracy.
A study suggests that smartwatch heart rate measurement algorithms are less effective in people with darker skin tones due to increased melanin absorption. Researchers emphasize the need for diverse population inclusion and explore alternative light wavelengths for more accurate readings.
A small study found statistically significant improvements in various symptoms, including fatigue and chest discomfort, after EECP therapy in long COVID patients. Researchers believe this is the most plausible explanation for the benefits derived from EECP and its link to long COVID.
Researchers at Wellesley College found that people frequently misinterpret bar graphs, with one in five graph readers categorically misunderstanding averages. The 'bar-tip limit error' occurs when viewers interpret the tip of the bar as the outer limit of the data, despite the average being the balanced center point.
A recent study has found that older adults who scored higher on a fatigue scale were more than twice as likely to die within 2.7 years compared to those with lower scores. Researchers say this can serve as an indicator for earlier mortality and encourage people to stick with exercise goals.
Researchers at Skoltech developed an algorithm that aggregates and maps environmental data to predict and visually map the quality of water, soil, or air. The solution uses interpolation to fill gaps in data and produces detailed maps with high resolution.
Researchers from Harvard University and others have found that large-scale COVID-19 vaccination surveys were off by up to 17 percentage points due to systematic biases in the data. The 'Big Data Paradox' highlights how big data sets can minimize one type of error while magnifying another, leading to misleading results.
Researchers developed a novel sequential analysis method to detect EWSs in daily COVID-19 case data across 24 countries, finding warnings were regularly detectable prior to exponential cases changes. The reliability of these signals depended on the time between successive waves and mathematical likelihood of critical transition.
A team of international researchers has developed a globally-applicable ethical code for ancient human DNA research, aiming to address social and cultural implications. The code proposes cooperation with stakeholders, minimal damage to human remains, and respect for indigenous communities' perspectives.
Researchers found that high water content in magma can reduce the risk of an explosive eruption. The study suggests combining different metrics to predict gas bubbles and crystallization, potentially leading to better forecasts for dangerous explosions.
Researchers used a supercomputer to emulate Google's quantum processor and discovered a reachability deficit, a performance limitation induced by a problem's constraint-to-variable ratio. The study showed that future experiments will require significantly more quantum resources to overcome this limit.
A new model developed by Stanford researchers provides a more accurate baseline for tracking methane emissions from oil and gas operations. The model, which uses bottom-up approach with component-level data, closely approximates top-down inventories and highlights specific activities that should be targeted for remediation or regulation.
A systematic analysis of probabilistic cost forecasts from expert-based and model-based methods reveals that current forecasts underestimates the pace of technological change. Model-based approaches outperformed expert elicitations in capturing observed costs, but both underestimated technological progress in most energy technologies.
Scientists recorded sound signals from March to June 2020 in Pennsylvania using underground fiber-optic sensors. They found a quiet period during the lockdown followed by a recovery of activity as restrictions lifted, with construction and vehicle traffic recovering sooner than pedestrian traffic.
The European Alliance of Associations for Rheumatology has developed guidelines for studies on work participation in people with inflammatory arthritis. The guidelines focus on identifying contextual factors and accounting for changes in work status over time.
Researchers at UT Austin developed a machine learning model that predicts the amount of lateral movement in soil during earthquakes, achieving 80% accuracy. The model uses over 7,000 data points from Christchurch, New Zealand, and was trained on the Frontera supercomputer.
A new study suggests that skipping breakfast leads to a nutritional gap, with adults who skip morning meals likely missing out on calcium, vitamin C, and fiber. The research analyzed data from over 30,000 American adults and found that those who skipped breakfast had lower intakes of essential nutrients for the entire day.
Groundwater extraction in Iran has declined by an average of 18% due to physical limits, while deep well extraction has increased, leading to further depletion. The study reveals widespread aquifer depletion and salinization across all 30 sub-basins, posing significant risks to the country's water resources.
Researchers at Michigan State University have developed a new dataset to diagnose broadband availability, revealing significant gaps in urban, suburban, and rural areas. The tool aims to help policymakers allocate resources efficiently to improve local conditions.
Margaret Martonosi's work on power-aware microarchitectures and narrow bit-widths has led to new fields of research. She has authored over 175 publications, including papers on dynamic thermal management and formal control-theoretic approaches for balancing power and performance.
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.
Computer scientists study the role of prior knowledge in machine learning tasks, considering philosophical frameworks that combine elements of rationalism and empiricism. This approach aims to improve the accuracy of complex systems by incorporating human-like reasoning and flexibility.
Researchers developed P-Flash, an AI-powered tool predicting flashover in burning buildings. It uses temperature data from heat detectors and shows promise in anticipating simulated flashovers, identifying unmodeled physical phenomena that can improve forecasting in real fires.
A new study published in Medical Care Research and Review found that Medicaid expansion provided a financial boost to hospitals, with uncompensated care costs declining by 28% in states that expanded Medicaid relative to 2013 levels. The study analyzed critical data at state and national levels using IRS and CMS data sources.
KICT has developed an effective structural monitoring technique to monitor massive infrastructures, such as long-span bridges. The method uses multi-fidelity data fusion to combine point and distributed strain sensors for accurate responses over whole infrastructures.
Researchers developed a 3D holographic head-up display using LiDAR data to create ultra-high-definition holographic representations of road objects, alerting drivers to potential hazards without distraction. The technology has shown promising results in early tests, suggesting it could be a valuable addition to existing safety measures.
Researchers found that severity of illness, history of stroke, and being divorced or widowed are predictive factors of delirium. Nearly 50% of patients admitted to the hospital had delirium at admission, highlighting the need for earlier identification and management.
A Texas A&M study found that racial and ethnic diversity in schools can decrease mental health risk for some Latinx students, while increasing symptoms among non-Latinx Black and Latinx students. The research used data from 389 sixth-graders across 14 urban Texas public schools.
Researchers at UMass Amherst used a novel approach to model ancient Antarctic ice sheets, revealing a thick but diminished ice sheet under mid-Miocene warmest conditions. The study suggests that greater precipitation led to a thickening of the ice sheet's interior regions, with potential implications for future sea level rise.
Researchers created a satellite-based map of human pressure on lands around the world using machine learning. The map reveals abrupt changes in landscapes due to deforestation, mining, and urbanization, providing insights into biodiversity conservation and sustainability.
Scientists develop a statistical randomness-based framework to optimally classify extremely large datasets. By analyzing the variance of sample sizes, they can identify the total number of attributes needed, providing a promising solution for high-performance computing and machine learning.
Researchers explore how blockchain and machine learning can optimize customer experiences across channels, while addressing data privacy concerns. The study identifies opportunities for future research on attribution modeling and incentivizing partner participation in blockchains.
Researchers studied 71 elephant seals' daily routines, finding they prioritize low-risk rest at night as they gain body fat. Seals with more fat prefer darker waters for safety over energy conservation, and adjust bedtime based on daylight hours.
Artist Michael McDermott creates an art exhibit using images, text, and audio from old cellular phones to raise awareness about digital data privacy. The project aims to educate viewers on the importance of safeguarding deleted information and best practices for disposing of old devices.
A new study reveals common seismological equipment is vulnerable to cyberattacks, posing risks to data collection and processing. Seismic networks may be exposed to security breaches if devices use non-encrypted data, insecure protocols, and poor user authentication mechanisms.