Research using machine learning to solve scientific problems. Covers AI quality control, sea-surface temperature estimation, label-free cell analysis, and cyber defense.
The Margot and Tom Pritzker Prize for AI in Science Research Excellence recognizes exceptional contributions to artificial intelligence and scientific inquiry. Awardees will receive $50,000 and be honored at the University of Chicago and Caltech Conference on AI+Science.
A new study from Regenstrief Institute presents a low-cost, scalable methodology for early dementia risk identification. The technique uses machine learning to extract relevant phrases from medical notes, providing healthcare providers with an individualized dementia risk prediction or evidence of mild cognitive impairment.
Researchers suggest that starting with smaller neural networks and using curriculum learning can improve performance and reduce the need for massive computing resources. This approach could lead to more resource-efficient and less energy-consuming AI systems.
Researchers developed a novel model predicting mood episodes using only sleep-wake pattern data from wearable devices. The study found daily changes in circadian rhythm are key predictors of mood episodes, offering new possibilities for tracking individual changes to prevent future episodes.
The discovery provides a unique way to investigate the extreme phase of stellar evolution, bridging the gap between the earliest and final stages of binary star systems. This breakthrough could help explain cosmic events like supernova explosions and gravitational waves.
A new machine learning algorithm reconstructs wind fields quickly and accurately, even with limited observational data. This enhances storm forecasting and hazard preparedness by providing valuable data on a tropical cyclone's intensity and potential impact.
DeltaAI enables scientists and researchers to address the world's most challenging problems by accelerating complex AI, machine learning, and high-performance computing applications. The system will quadruple NCSA's AI-focused computing capacity and expand the NSF-funded advanced computing ecosystem.
Scientists have developed a methodology that can predict the most favorable binding sites of gold nanoparticles to five common human blood proteins. This breakthrough enables researchers to investigate how drug-carrying nanoparticles interact with blood proteins, which could lead to more effective cancer treatments.
Researchers found that membership inference attacks on large language models (LLMs) are not effective in measuring information exposure risks. The common method used to test LLM leaks suffers from ambiguity due to the fluidity of language, making it difficult to define a representative set of non-member candidates.
A WVU research partnership with the DEA aims to improve fast and accurate identification of psychoactive substances like fentanyl. The Expert Algorithm for Substance Identification (EASI) will enable labs using different instruments to share data on chemical profiles, helping to identify drugs like fentanyl.
Scientists use machine learning-based classifiers to differentiate lung cancer and noncancer based on urinary miRNA ensembles. The study reveals high specificity and sensitivity in detecting early-stage lung cancer, offering new hope for improved patient outcomes.
Neuro-oncology experts developed guidelines to standardize AI use in brain cancer diagnosis and treatment. The recommendations aim to ensure reliable clinical trial results and protect patients.
A new study developed a machine-learning model to predict microbial load, the density of microbes in our guts. The model revealed that many factors can influence microbial load, including lifestyle, diseases, and medications.
A recent machine learning study suggests that the association between gut bacteria and disease may be overstated. Instead, changes in microbial load were found to be a key factor in the presence of disease-associated microbial species. This discovery challenges current understanding of the gut microbiome's role in disease etiology.
Researchers developed a robot that identifies plants by measuring leaf properties with an electrode, achieving an average accuracy of 97.7% for ten different species. The device may revolutionize crop management and early disease detection, but its limitations need to be addressed.
The partnership aims to improve the speed and accuracy of flood damage assessments by combining multi-modal data sources and a novel vision language model. The system will enable first responders to quickly allocate resources and respond more efficiently to weather disasters.
Researchers developed a machine learning framework that predicts inorganic pollutants in groundwater based on limited water quality samples. The model suggests 15% to 55% of sites may truly be risk-free, identifying critical gaps in groundwater quality datasets.
A group of student researchers developed a machine learning-based approach to better understand and represent the decline in global ocean oxygen levels. Their research showed that the world's oceans have lost oxygen at a rate of about 0.7% per decade from 1970 to 2010.
NeuroMechFly v2 simulates how a fruit fly navigates through its environment while reacting to sights, smells, and obstacles. The model can track moving objects visually or navigate towards an odor source, while avoiding obstacles in its path, enabling researchers to study brain-body coordination and animal intelligence.
HemaChrome's machine learning-based technology enables instant and noninvasive measurement of blood hemoglobin levels from digital photos, facilitating point-of-care diagnostic tests. The collaboration with Global Health Labs aims to address anemia diagnosis gaps in low- and middle-income countries.
Researchers used machine learning to analyze data from 50 CF-LVAD patients and identified six prognostic factors that predicted post-implantation stroke. Lower levels of OxPhos proteins were associated with an increased risk of new strokes after implantation.
The university introduces Bachelor of Engineering in Robotics and Double Major Bachelor of Engineering Science in Process Engineering and Synthetic Chemistry to address global demand for roboticists. The programs incorporate AI-related elements and are designed to provide students with a unique skill set.
Researchers at Lehigh University are using advanced algorithms and cross-domain data to help cities predict human movement patterns, enabling better planning and preparedness for events and emergencies. The model will account for variations in data streams from different sources, such as cell towers, GPS, and financial transactions.
Researchers developed innovative encoding methods that simplified quantum circuits for data encoding, reducing circuit depth by a factor of 100 while maintaining accuracy. These methods showed improved resilience against adversarial attacks, paving the way for practical application of quantum machine learning on current devices.
Researchers identify biomarker Ki67 that indicates whether treatment with vedolizumab will be successful. The biomarker can help predict which patients are more likely to respond to the form of treatment, allowing for more targeted use and potentially improving patient outcomes.
A TU Wien-developed robot can learn to clean a sink by watching humans perform the task, adapting its knowledge to different shapes and applying the right amount of force. The technology combines machine learning and robotics, enabling robots to share their parameters through federated learning.
A team of scientists leveraged machine learning to find promising compositions for sodium-ion batteries, achieving exceptional energy density. The study trained a model on a database of 100 samples to predict the optimal ratio of elements needed to balance properties like operating voltage and capacity retention.
The University of Tennessee and Lockheed Martin have expanded their master research agreement to address national security challenges. The partnership will support advanced technologies such as hypersonics, materials, and energy systems, while also providing experiential learning opportunities for students.
Researchers have identified two genes, ATXN2L and MMP14, linked to both rheumatoid arthritis and osteoporosis. These genes play a role in apoptosis, immune regulation, and bone metabolism.
The AI-powered system can detect toxic gases like nitrogen dioxide in real-time, identifying the source of harmful gas leaks. The system's optimization technique ensures fewer resources are used while providing faster and more accurate gas leak detection.
The project aims to identify and fabricate optimized first-wall materials using advanced computer simulations enhanced by machine learning, accelerating the discovery of new materials by 100-fold. The research will leverage synthesis, irradiation, and testing facilities to conduct a high-impact materials discovery campaign.
Researchers developed AI to predict bloodstream infections and antimicrobial resistance in ICU patients, providing same-day assessments. This technology is cost-effective and faster than current methods, enabling quicker decision-making on antibiotic use.
The São Paulo School of Advanced Science on High-Dimensional Modeling offers minicourses and sessions to enhance data professionals' training in machine learning and finance. Key challenges related to forecasting, asset allocation, and climate econometrics will be addressed through state-of-the-art science and research.
Researchers developed a wearable ultrasound device that tracks muscle function without invasive procedures, offering high-resolution imaging and wireless monitoring capabilities. The technology has potential applications in respiratory health and human-machine interfaces.
Srisharan Shreedharan leads a collaborative effort to gain knowledge of processes that could improve seismic hazard forecasting. The research aims to identify key indicators that can help predict earthquakes and reduce seismic hazards.
Yihao Zheng and his team are developing a fiber-optic probe that analyzes artery blockages in the brain and guides procedures for blockage removal. The technology uses light and advanced calculations to determine the properties of blood clots, enabling doctors to make informed decisions about how to remove them.
A new study published in Nature found that up to 215 million hectares of land in tropical regions around the world has the potential to naturally regrow, storing 23.4 gigatons of carbon over 30 years. The study identified areas with high regrowth potential based on factors such as soil quality and proximity to existing forest.
A research team from Hokkaido University has developed a flexible multimodal wearable sensor patch that can detect arrhythmia, coughs, and falls using edge computing on a smartphone. The sensor patch generates large amounts of data that must be processed to be understood.
Researchers found that Black patients are less likely to receive medical tests, leading to inaccurate AI models. A new algorithm corrects for this bias by identifying untested patients based on race and vital signs, improving model accuracy to around 60%.
Researchers analyzed correlations between fruit fly and human data to identify key metabolites impacting lifespan. Threonine was found to extend lifespan in flies and show promise as a therapeutic target for aging interventions.
A new study by UCLA Health reveals that standard medical record surveillance methods miss youth with suicidal thoughts and behaviors in children, boys, and Black and Hispanic youths. Machine learning algorithms improved detection rates when incorporating additional data from visit notes.
A machine learning model predicts soil behavior during earthquakes, identifying areas vulnerable to liquefaction and providing contour maps for safer construction sites. The study uses geological data to create detailed 3D maps of soil layers, improving prediction accuracy by 20%.
Researchers have identified a genetic signature that can predict neonatal sepsis in newborns before symptoms appear, allowing for earlier recognition and life-saving treatment. The discovery has the potential to improve healthcare outcomes in lower- and middle-income countries where neonatal sepsis is most prevalent.
A machine learning algorithm developed by University of Cambridge researchers can detect and grade heart murmurs in dogs with high accuracy, similar to expert cardiologists. The technology has the potential to empower primary care veterinarians to provide early detection and treatment, improving quality of life for dogs.
Researchers developed a machine learning model to predict dielectric function of materials, facilitating novel dielectric material development. The model speeds up calculations by using chemical bonds between atoms and achieving accuracy close to first-principle calculations.
A new study uses PCA and Mahalanobis distance to detect early bolt loosening in wind turbines, achieving accuracy of over 95%. This technology improves operational safety and efficiency by providing early warnings based on objective data.
Researchers at UC Santa Cruz have developed a highly accurate and affordable spectrometer that can be customized for specific applications. The device uses machine learning algorithms to reconstruct images with high accuracy, enabling astronomers to study phenomena such as exoplanet atmospheres and dark matter in faint galaxies.
A new study by Penn State researchers suggests that making AI training data diversity information available can shape users' expectations of algorithmic fairness and trust. Displaying racial diversity cues in AI interfaces can enhance users' perceptions of algorithmic fairness and trust, according to the study's findings.
A new tool, SymGen, enables users to verify AI model responses more quickly and easily by displaying data citations. This speeds up the manual validation process by 20 percent, making it easier for users to spot errors in LLMs deployed in various real-world situations.
A new AI-powered model has been developed to predict kidney transplant outcomes with high accuracy, offering hope for more efficient organ allocation and improved patient outcomes. The tool, UK-DTOP, outperforms existing methods in predicting outcomes for deceased-donor kidney transplants.
A recent study reviews advancements in reinforcement learning for autonomous vehicle control, highlighting similarities and differences in DRL formulations and training algorithms. The research aims to enhance RL applications, making autonomous vehicles more capable of handling complex traffic situations under uncertain conditions.
The improved method achieves high accuracy in lithium-ion battery state of charge estimation, outperforming traditional methods such as Back propagation Neural Network and Long Short-Term Memory. The model's robustness is enhanced through periodic parameter updates based on battery operating conditions.
Researchers have created a new diagnostic tool using machine learning to detect schistosomiasis, a persistent parasitic infection affecting an estimated 250 million people. The tool can identify low levels of the infection and distinguish between active and past infections, leading to earlier treatment and improved long-term outcomes.
Researchers developed an AI model that uses machine learning and combination theory to predict suspicious skin lesions. The new C4C Risk Score has an accuracy of 69% and significantly outperformed existing methods, including 7PCL and Williams score.
A new project, 'Crowd-Assisted Human-AI Teaming with Explanations,' aims to develop an interactive AI system that leverages the collective strengths of human crowd workers and machine learning models. The researchers will use crowdsourcing platforms to recruit experts and non-experts to perform tasks, making the system more robust and ...
The new center aims to develop AI-driven tools for predicting solar eruptions, expand space science education programs, and build a long-term dataset of sun activity. It will also establish an education program providing research opportunities for students and promoting STEM education.
A study from Lehigh University and Seattle University found that making patients aware of biases in human healthcare decisions increases receptiveness to AI recommendations. By highlighting the limitations of human judgment, healthcare providers can create a more balanced relationship between patients and emerging technologies.
Researchers use machine learning to analyze optimal bike lane placement in Toronto, balancing accessibility for all with overall efficiency. Key findings include a trade-off between equity and utility, with essential routes like Bloor West's bike lanes serving neighbourhoods far from their endpoints.
Researchers have developed a new system for full-body motion capture that leverages sensors within consumer mobile devices. The app, called MobilePoser, tracks a person's full-body pose and global translation in space in real time with advanced machine learning and physics-based optimization.
Computer simulations point the way towards better solar cells by gaining crucial insights into what influences properties of 2D perovskite materials. Researchers have discovered that the choice of organic linkers can directly control how atoms in surface layers move, affecting optical properties.