Science – Page 120 – Minority Mindset

Beyond Connections: Predicting Trust in Evolving Networks

15.12.2025 by qfx

The study demonstrates a context-aware trust prediction system’s ability to differentiate between user pairs-accurately identifying distrusted relationships in some cases and trusted relationships in others-suggesting the model captures nuanced interactions beyond simple overall assessments.

A new model leverages contextual understanding and dynamic graph analysis to anticipate reliable relationships within complex systems.

Seeing Through the Blockchain: AI Detects Smart Contract Attacks

15.12.2025 by qfx

A novel model is proposed for detecting cyberattacks targeting Ethereum smart contracts and transactions.

A new approach uses computer vision techniques to analyze blockchain transactions and identify malicious activity within smart contracts.

When AI Minds Get Fuzzy: The Limits of Clinical Reasoning

15.12.2025 by qfx

New research reveals that artificial intelligence, despite vast medical knowledge, struggles to reliably interpret complex, real-world patient data.

Boston’s Green Line: Mapping Resilience in an Urban Rail Network

15.12.2025 by qfx

The network’s data, when subjected to risk calculation, demonstrates an inevitable trajectory toward decay, quantified as a function of systemic vulnerability rather than simply elapsed time.

A new analysis details the vulnerabilities and strengths of Boston’s Green Line subway system, offering critical insights for improving its operational reliability and security.

Predictable Robots: Calibrating Uncertainty for Safer Action

15.12.2025 by qfx

$The system leverages state transition data alongside an approximate dynamical model to establish a symmetry-aware probabilistic error bound on configuration predictions, enabling the computation of a calibrated prediction region $𝒞^q$ guaranteed to contain the true configuration resulting from a desired action $u_{des}$.$

A new approach uses advanced mathematical tools to build more reliable predictions of robot movements, even with imperfect information about the environment.

AI Guides Heart Valve Treatment Choices

14.12.2025 by qfx

Patients undergoing transcatheter aortic valve replacement (TAVR) disproportionately represent higher-risk profiles within both the Hartford and St. Vincent’s cohorts, necessitating prognostic matching on training data to mitigate treatment imbalance and ensure reliable comparative analysis.

A new data-driven tool helps clinicians determine the best treatment – surgical or transcatheter – for patients with aortic stenosis.

Future-Proofing the Farm: Planning for Uncertainty in Crop Production

14.12.2025 by qfx

The proposed framework generates a spatio-temporal crop allocation plan for the year 2024, demonstrating a system capable of dynamic resource distribution across both space and time.

A new framework combines economic viability with sustainable agricultural practices to create resilient crop plans in the face of unpredictable conditions.

When AI Forgets What It Knows: Adapting to New Realities

14.12.2025 by qfx

$During test-time adaptation, a novel approach blends geometric proximity with predictive confidence to create hybrid targets, subtly guiding features toward plausible classifier weights while actively repelling negatives through the loss function $\mathcal{L}\_{\text{NC}}$.$

A new study reveals how neural networks lose accuracy when faced with unfamiliar data, and introduces a method to realign their internal representations for better performance.

From Text to Knowledge: A Smarter Way to Classify Data

14.12.2025 by qfx

The pipeline transforms textual documents into a graph structure, initially extracting representations and defining relationships between nodes, then leverages a large language model to label a selection of these nodes before propagating those labels across the entire graph using a graph neural network.

A new approach combines the power of language models with graph networks to achieve accurate text classification, even when labeled data is limited.

Fast and Accurate Aerodynamic Prediction with Kernelized Neural Surrogates

14.12.2025 by qfx

A new approach combines the strengths of kernel methods and neural networks to dramatically accelerate aerodynamic simulations while maintaining high accuracy.