Machine learning maps fifteen years of London crime, street by street
An open-access study of nearly seventeen million Metropolitan Police records shows an interpretable decision tree modestly outperforms naive averages at ...
Tag: Machine Learning
From Fourier to Deep Learning: How Bridge Health Monitoring Signals Are Evolving
A new review of 201 studies traces how bridge structural health monitoring evolved from classical Fourier and wavelet analysis to ...
Eleven Papers Map the Frontier of Data Spaces and Data Governance
A new special issue of Data Science and Engineering collects eleven peer-reviewed papers on data spaces, data governance, and data-driven ...
Low serum magnesium linked to early sepsis-related kidney injury risk
A retrospective machine learning study of 19,805 ICU patients with sepsis found that lower early serum magnesium levels were consistently ...
Brain Signal Patterns Shift as Freezing of Gait Strikes Parkinson’s Patients
A hemispheric-gradient EEG microstate and muscle activity both dip during freezing of gait in Parkinson's disease, and EMG features can ...
Machine Learning Outperforms Design Codes at Predicting Concrete Crack Spacing
Extremely randomized trees models trained on 96 experimental records predicted primary crack spacing in reinforced concrete flexural members more accurately ...
Diffusion Model Turns Sparse User Histories into Better Cold-Start Recommendations
A new diffusion-based framework extracts transferable preference patterns from historical user data to improve recommendations for users with almost no ...
VIPR Software Framework Brings Machine Learning to Scientific Inverse Problems
The open-source VIPR framework standardizes machine-learning inference for ill-posed inverse problems in reflectometry and beyond through configurable, reproducible workflows.
Speech Alone Predicts Schizophrenia Symptom Severity Across Ten Countries
An international study shows machine learning models can estimate schizophrenia symptom severity from speech recordings alone across ten languages.
Digital Twin Falls Short: Physics-Based Blood Pressure Model Fails Rigorous Stress Test
A physics-constrained arterial-pressure twin candidate built from photoplethysmography showed no reconstruction benefit, weak forecasting, and uninformative uncertainty under locked, subject-disjoint ...
Graph Neural Network Maps Local Brain Aging Across the Cortex in Health and Alzheimer’s Disease
A graph neural network trained on more than 14,000 brain scans maps local cortical aging at millimeter resolution and links ...
Deep learning reads well logs to map coal-bearing rock types, with simple CNN outperforming complex models
A new study combined SEM, XRD and NMR pore characterization with six machine-learning models on 37,227 well-log records, finding that ...
AI foundation model pretrained on 500 million collision events transforms particle physics analysis
Researchers introduced EVENET, a foundation model pretrained on 500 million simulated collision events that outperforms task-specific baselines across new-physics searches, ...
New BASM framework charts a maturity path for AI in clinical biochemistry labs
A new open-access framework called BASM maps five maturity tiers for AI-augmented clinical biochemistry, linking analytics, algorithmic governance, and health-service ...
Research report: CLeaR: An adaptive continual learning framework for regression tasks
Researchers have developed CLeaR, a continual learning framework that lets neural networks handle regression tasks such as wind power forecasting ...
Scoping review maps CT evidence on sinuses for sex, age, stature estimation
A scoping review of 76 studies from 18 countries finds CT-based sinus and craniofacial measurements are well studied for sex ...
Machine Learning Calculator Estimates One-Year Mortality Risk Before Joint Replacement
A machine learning model trained on over 234,000 joint replacement records estimates one-year postoperative mortality risk from routinely available preoperative ...
Smarter Whale Algorithm Sharpens AI Predictions of Ultra-High-Performance Concrete Strength
Researchers combined a strengthened whale optimization algorithm with a kernel extreme learning machine to predict ultra-high-performance concrete strength, achieving test ...
Simple Machine Learning Outperforms Deep Models in Somali Dialect Test
A new study reports that a simple Naive Bayes classifier reached 99.05 percent accuracy in distinguishing Somali's MAXAA TIRI and ...
Blockchain Shield Uses Smart Contracts to Block DDoS Attacks on Servers
A blockchain-based framework using Ethereum smart contracts and a CNN detector blocked 99.8% of simulated DDoS attacks on client–server networks ...
