Results 81 to 90 of about 29,456 (261)

Trustworthy Multimodal Attention Framework for Creep Rupture Life Prediction Under Data‐Scarce Conditions: A Case Study on IN718

Advanced Science, EarlyView.
An attention‐based multimodal deep learning framework is developed to predict the creep life of Ni‐based superalloys by fusing processing parameters with microstructural micrographs. The model achieves high accuracy (R2 = 0.92), aligns with metallurgical principles by capturing δ‐phase influence, and incorporates uncertainty quantification, offering a ...
Haopeng Lv, Dayong Wu, Ziyuan Rao, Qian Wang, Haikun Ma, Jie Kang, Wang Li, Huicong Dong, Yandong Wang, Zhinan Yang, Ru Su   +10 more
wiley   +1 more source

Biochemically Constrained Multi‐Omics Integration Reveals Protein–Metabolite Dependencies Across Diseases

Advanced Science, EarlyView.
ProMetNet introduces a biologically constrained deep learning framework for proteo‐metabolomic integration by embedding Reactome‐derived pathway topology into neural networks. It captures non‐linear molecular dependencies and pathway‐level metabolic reorganization, enabling interpretable discrimination.
Minghui Zhao, Na Zhou, Ruotong Liu, Xiaofei Li, Jian Li, Fuzhong Xue, Qingzhen Hou   +6 more
wiley   +1 more source

NICE: A Two‐Step Non‐Invasive Framework for Embryo cfDNA Read Enrichment and Quality Assessment

Advanced Science, EarlyView.
The non‐invasive NICE framework, built on an ensemble stacking machine learning model, prioritizes embryos by analyzing cell‐free DNA from spent culture medium. By integrating multimodal signals, including genomic and epigenetic profiles, this automated approach standardizes morphological assessment without human bias, paving the way for more precise ...
Xueya Zhou, Shu Ding, Zhenyi Zhang, Qiaoling Shangguan, Jie Qiao, Peijie Zhou, Yidong Chen   +6 more
wiley   +1 more source

Adversarial Evasion Attacks on Computer Vision using SHAP Values

The paper introduces a white-box attack on computer vision models using SHAP values. It demonstrates how adversarial evasion attacks can compromise the performance of deep learning models by reducing output confidence or inducing misclassifications. Such attacks are particularly insidious as they can deceive the perception of an algorithm while eluding
Frank Mollard, Marcus Becker, Florian Roehrbein   +2 more
openaire   +2 more sources

Quantum‐Trained AI Enables Inverse Design of Organic Frontier Orbitals at Billion‐Scale

Advanced Science, EarlyView.
Quantum‐trained AI links molecular substructures to frontier‐orbital energetics and enables interpretable screening across nearly one billion GDB‐13 molecules. By combining fragment‐ and ring‐level insights with donor–acceptor energy alignment against ITIC, the framework narrows an immense chemical space to a small set of promising candidates and ...
Yeongnam Ko, Se Jin Kim, Ki Chul Kim
wiley   +1 more source

CQD-SHAP: Explainable Complex Query Answering via Shapley Values

CoRR
Complex query answering (CQA) goes beyond the widely studied link prediction task by addressing more sophisticated queries that require multi-hop reasoning over incomplete knowledge graphs (KGs). Research on neural and neurosymbolic CQA methods is still an emerging field.
Parsa Abbasi, Stefan Heindorf
openaire   +2 more sources

Robots and Minimal, Physics‐Informed Features: A Hybrid Framework for Enzyme Catalysis

Advanced Science, EarlyView.
Robotic experimentation and physics‐informed machine learning combine to predict enzyme substrate scope. With a handful of interpretable features derived from docking and quantum mechanics calculations, our model rivals descriptor‐heavy AI approaches and extrapolates to unseen substrates and enzyme classes.
Natalia Onishchenko, Eric S. Larsen, Govind Paneru, Kisung Lee, Diana V. Kolygina, Yankai Jia, Elizabeth Maria Clarissa, Wai‐Shing Wong, Bartosz A. Grzybowski   +8 more
wiley   +1 more source

Feature selection strategies: a comparative analysis of SHAP-value and importance-based methods

Journal of Big Data
In the context of high-dimensional credit card fraud data, researchers and practitioners commonly utilize feature selection techniques to enhance the performance of fraud detection models.
Huanjing Wang, Qianxin Liang, John T. Hancock, Taghi M. Khoshgoftaar   +3 more
doaj   +1 more source

Hypertension control in resource-constrained settings: Bridging socioeconomic gaps with predictive insights

International Journal of Cardiology. Cardiovascular Risk and Prevention
Background:: Hypertension continues to be a pivotal driver of global cardiovascular disease burden and adverse health outcomes, particularly in resource-constrained settings where disparities in socioeconomic status and clinical infrastructure hinder ...
Md Abul Kalam Azad, Md Abu Sufian, Lujain Alsadder, Sadia Zaman, Wahiba Hamzi, Amira Ali, Md. Zakir Hossain, Boumediene Hamzi   +7 more
doaj   +1 more source

Proteogenomic Profiling of Idiopathic Pulmonary Arterial Hypertension Identifies Sex‐Differential Proteins and Candidate Therapeutic Targets

Advanced Science, EarlyView.
An integrated proteogenomic analysis of 44,137 predominantly European‐ancestry UK Biobank participants aged 40–69 years identifies 12 robust proteins associated with idiopathic pulmonary arterial hypertension. These proteins define a high‐mortality molecular endotype, support early detection and mortality prediction, reveal sex‐differential proteomic ...
Xinjie Lin, Yangchang Zhang, Yang Pu, Duanke Liu, Tao Lei, Kai Ma, Qiyu He, Zheng Dou, Yuze Liu, Dengyuan Liu, Yinge He, Yanshang Wang, Xiaojiao Zheng, Yanbing Ma, Jianrong Zhou, Weiding Zhai, Binbin Su, Li Chen, Shoujun Li   +18 more
wiley   +1 more source