A5100310779- High-Temperature Coating Behaviors
- Machine Learning in Healthcare
- Topic Modeling
- Metal Alloys Wear and Properties
- Adversarial Robustness in Machine Learning
- Materials Engineering and Processing
- Aluminum Alloys Composites Properties
- Advanced ceramic materials synthesis
- Artificial Intelligence in Healthcare and Education
- Drilling and Well Engineering
- Privacy-Preserving Technologies in Data
- Natural Language Processing Techniques
- Mineral Processing and Grinding
- Artificial Intelligence in Healthcare
- Anomaly Detection Techniques and Applications
- Digital Radiography and Breast Imaging
- Fault Detection and Control Systems
- Peer-to-Peer Network Technologies
- Integrated Circuits and Semiconductor Failure Analysis
- Advanced materials and composites
- Magnetic and Electromagnetic Effects
- Domain Adaptation and Few-Shot Learning
- Grey System Theory Applications
- Plant responses to water stress
- Intracerebral and Subarachnoid Hemorrhage Research
National Synchrotron Radiation Laboratory
2024-2025
University of Science and Technology of China
2024-2025
Chinese University of Hong Kong
2023-2025
Pennsylvania State University
2023-2024
China University of Geosciences
2024
Chinese Academy of Sciences
2024
Hebei Medical University
2024
Hunan University
2023-2024
Sun Yat-sen University
2024
Nanjing University of Information Science and Technology
2023
Abstract The production of formic acid via electrochemical CO 2 reduction may serve as a key link for the carbon cycle in economy, yet its practical feasibility is largely limited by quantity and concentration product. Here we demonstrate continuous at M an industrial‐level current densities (i.e., 200 mA cm −2 ) 300 h on membrane electrode assembly using scalable lattice‐distorted bismuth catalysts. optimized catalysts also enable Faradaic efficiency formate 94.2 % highest partial density...
A novel ternary eutectic salt, NaNO3-KNO3-Na2SO4 (TMS), was designed and prepared for thermal energy storage (TES) to address the issues of narrow temperature range low specific heat solar salt molten salt. The thermo-physical properties TMS-2, such as melting point, decomposition temperature, fusion enthalpy, density, viscosity, capacity volumetric (ETES), were determined. Furthermore, a comparison between commercial TMS-2 carried out. had point 6.5 °C lower 38.93 higher than those use TMS...
The development of electronic health records (EHR) systems has enabled the collection a vast amount digitized patient data. However, utilizing EHR data for predictive modeling presents several challenges due to its unique characteristics. With advancements in machine learning techniques, deep demonstrated superiority various applications, including healthcare. This survey systematically reviews recent advances learning-based models using Specifically, we introduce background and provide...
Maximizing the utilization of in situ extraterrestrial resources, including solar-powered water electrolysis using lunar soil as a catalyst, is promising strategy for achieving sustainable fuel and oxygen supply exploration. However, these soil-based silicate minerals suffer from unsatisfactory intrinsic activity splitting due to poor electrical conductivity lack catalytic sites. Here we report use simple Joule-heating method sinter into disordered matrix at ∼2000 °C. The as-prepared...
Thyroid-associated orbitopathy (TAO) is a prevalent inflammatory autoimmune disorder, leading to orbital disfigurement and visual disability. Automatic comprehensive segmentation tailored for quantitative multi-modal MRI assessment of TAO holds enormous promise but still lacking. In this paper, we propose novel method, named cross-modal attentive self-training (CMAST), the multi-organ in using partially labeled unaligned data. Our method first introduces dedicatedly designed pseudo label...
Synthesizing electronic health records (EHR) is essential for addressing data scarcity, bias, and fairness in healthcare models. EHR are inherently multimodal sequential, encompassing structured codes, clinical notes, medical images, irregular time intervals. Traditional generative models like GANs VAEs struggle to capture these complexities, while diffusion-based offer improvements but remain limited task-specific applications. To address challenges, two models, MedDiffusion EHRPD, have...
To explore the Co/Cr ratio impact on high-velocity oxygen fuel (HVOF)-sprayed WC-Co-Cr coatings microstructure and performances, three kinds of coatings, namely WC-4Co-10Cr, WC-7Co-7Cr, WC-10Co-4Cr, were prepared by using a (HVOF) spraying process. The coatings’ phase composition, microstructure, basic mechanical properties, abrasive wear, corrosion resistance investigated. results show that all comprise main WC, minor W2C, amorphous W-Co-Cr phase, besides WC-4Co-10Cr coating containing...
Terahertz (THz) plasmonic resonance, particularly on artificially designed metasurfaces, has garnered considerable attention for enhancing the identification of biomolecules. However, narrow-band response these resonances hinders comprehensive analysis sample characteristic fingerprints, leading to false-positive results. Here, we developed a multifunctional THz biosensor based broad-spectrum multi-Fano resonance capable noninvasive and ultrasensitive detection trace biological molecular...
This study proposes a short-term load prediction method of bidirectional long memory network based on feature mining the power consumption big data in combination with attention mechanism (AT) Bayesian optimization to address problems that considerable amount factors exist and relationship is obscured historical data. The comprehensively considers global features space local time. First, Cen-CK-means clustering used cluster electricity users, statistical, combination, time category...
Pretraining has proven to be a powerful technique in natural language processing (NLP), exhibiting remarkable success various NLP downstream tasks. However, the medical domain, existing pretrained models on electronic health records (EHR) fail capture hierarchical nature of EHR data, limiting their generalization capability across diverse tasks using single model. To tackle this challenge, paper introduces novel, general, and unified pretraining framework called MedHMP, specifically designed...
Summary High-voltage electric pulse rock-breaking (HVEPB) has proved to be a novel and inexpensive method of breaking rock regardless composition, but the design electrode drill bit lacks theoretical basis. In this paper, we first establish plasma channel model for breakdown numerical HVEPB, which can simulate electrical effect different bits on HVEPB. Second, analyze effects arrangement structures high-voltage angles channels internal cracks processes through simulation. Finally, describe...
The widespread adoption of Electronic Health Record (EHR) systems in healthcare institutes has generated vast amounts medical data, offering significant opportunities for improving services through deep learning techniques. However, the complex and diverse modalities feature structures real-world EHR data pose great challenges model design. To address multi-modality challenge current approaches primarily rely on hand-crafted architectures based intuition empirical experiences, leading to...
Although pre-training has become a prevalent approach for addressing various biomedical tasks, the current efficacy of pre-trained models is hindered by their reliance on limited scope medical sources. This limitation results in data scarcity during and restricts range applicable downstream tasks. In response to these challenges, we develop Medical Cross-Source Pre-training (MEDCSP), new strategy designed bridge gap between multimodal MEDCSP employs modality-level aggregation unify patient...
A novelty method of 2DGabor-KDA(kernel Fisher discriminant analysis) for face recognition is proposed. First all, every facial image segmented into several sub-areas according to the five particular parts and then features key are extracted through 2DGabor wavelet, average values calculated from feature vectors gained corresponding pixel each test sample eigenvectors gained, in next place, KDA applied kernel-process eigenvectors, SVM(Support Vector Machine) adopted recognize images. The...
As a core component in the gas transmission process, internal wall surface of pressure regulator is prone to failure due long-term exposure high-pressure environment, resulting poor reliability regulator. Thus, thermo-hydro-mechanical coupling model for FL established this paper, and results are verified by engineering data. The effect valve opening on parameters (temperature, deformation, stress) studied detail through simulation. show that stress greater at sleeve, bore, outlet seat under...
The vulnerability of deep neural networks to adversarial samples poses significant security concerns. Previous empirical analyses have shown that increasing robustness through training leads models making unconfident decisions, thereby undermining trust in model confidence scores as an accurate indication their reliability. This raises the question: are and calibration mutually exclusive? In this work, we find empirically not only do examples mislead undefended make more confident mistakes...