AI Model Predicts Secondary-School English Performance With Built-In Uncertainty
A new computational framework claims the best mean predictive performance across four English learner corpora, forecasting writing scores and speaking-proficiency ...
Tag: Machine Learning
New Federated Learning Aggregation Rules Keep AI Training Stable When Client Data Diverge
Researchers propose CROG and CLIG, two aggregation algorithms that weight client updates by Euclidean gradient distances to stabilize federated learning ...
AlphaFold 3 Structures Can Guide T-Cell Receptor Engineering, With Caveats
A benchmark of 438 TCR–pMHC mutations shows experimental structures remain the best input for binding-affinity predictors, while AlphaFold 3 mutant ...
Research report: Metabolic signatures and machine learning identify gut-liver-heart axis dysfunction as a potential link to major adverse cardiovascular events in coronary artery disease
A retrospective metabolomics study in Hong Kong linked oxidized linoleic acid metabolites, altered PPAR signaling, and shifted bile acid-FGF19 levels ...
New Model Predicts How Much Strength Fire-Damaged Concrete Can Regain
A statistical and machine learning study of 464 experimental datasets yields an empirical model estimating how much compressive strength fire-damaged ...
Gut Microbes May Decide Whether Your Medicine Works or Harms You
A new review explains how gut microbial enzymes, bioaccumulation, and enterohepatic recycling can activate, inactivate, or toxify drugs, while cautioning ...
Frailty, not age alone, shapes recovery after stroke, editorial argues
A Frontiers in Neurology editorial argues that frailty, rather than chronological age alone, should guide stroke care decisions as populations ...
Graph neural network method boosts detection of rare cyberattacks in imbalanced data
A new graph neural network framework called EGNN expands rare attack samples and uses threshold-controlled pseudo-labeling to improve intrusion detection ...
Systematic Review Weighs Which Machine Learning Models Best Detect Bias in Mortgage Lending
A systematic review of mortgage-lending fairness research finds decision trees offer the most interpretable bias detection while neural networks and ...
AI Framework Mines Traditional Chinese Medicine for Phenotype-Based Drug Discovery
An AI framework combining Kolmogorov–Arnold network drug–target prediction with AutoML modeling of Chinese Materia Medica properties identified lipid-lowering candidates from ...
New host telemetry dataset enables robust and explainable ransomware detection and family attribution
Researchers built a hybrid host-level dataset of 5,326 ransomware and goodware executions with 392 features, achieving F1-scores of 0.982 for ...
Machine Learning Pinpoints Seven Genes and Key Liver Cells Behind MASH Fibrosis
An integrative RNA sequencing and machine learning study identified a seven-gene signature and key liver cell subpopulations associated with advanced ...
Biochar May Boost Wastewater Nitrogen Removal by Bridging Microbial Electron Transfer
A new review proposes three mechanisms by which biochar may act as an electron bridge to enhance anammox-based nitrogen removal ...
Machine Learning Ties Spinal Cord Scores to Everyday Independence After Injury
Machine learning models classified grooming, bladder management, and indoor mobility after spinal cord injury from standardized neurological scores, with accuracy ...
Tissue marker KIM-1 flags early kidney cancer recurrence missed by adjuvant criteria
In a two-centre exploratory study of 117 clear-cell renal cell carcinoma patients, 43.5% of recurrences within 24 months occurred among ...
AI Moves From Lab to Front Line of Outbreak Response, WHO Forum Hears
A WHO forum report describes how AI tools for outbreak forecasting, genomic surveillance and decision support are advancing, while experts ...
Machine learning ranks building damage after Nepal’s deadly 2015 Gorkha earthquake
A study of 762,106 buildings damaged in Nepal's 2015 Gorkha earthquake combines an MLP classifier with imbalance correction, sensitivity analysis, ...
New AI tool annotates single-cell data with far fewer labels
A new dual-view machine learning framework called scTrans-SVD achieved a mean Macro-F1 of 0.8820 across 13 public single-cell transcriptomic datasets ...
Blood metabolites may flag who will develop Chagas heart disease
Metabolomics analysis identified serum small molecule panels that predict progression from asymptomatic Chagas infection to cardiac disease.
Machine Learning Framework Maps a Closed-Loop Path to Defect-Free Polymer 3D Printing
A new review of 75 studies proposes a unified data-driven framework linking polymer 3D printing defects, multi-sensor monitoring, and machine ...
