Latest Articles for Graph Neural Networks

A new approach to reduce oversmoothing in transformers and GNNs.

New method improves machine learning performance across different data environments.

This study presents a new method for augmenting graphs using randomized Schur complements.

New machine learning models improve predictions in fluid dynamics.

DASH offers a fast and accurate approach for assigning partial charges in molecules.

A new method improves how GNNs explain their predictions.

Research shows how GNNs can adapt across different graph sizes.

Researchers develop a model to better predict protein stability changes from amino acid mutations.

A method to enhance MLPs using reliable knowledge from GNNs.

This paper examines how local homophily affects graph neural network performance.

A novel method enhances graph classification results using unsupervised domain adaptation.

This study examines AI's role in solving Hamiltonian cycles more efficiently.

A new approach emphasizes the importance of relationships in tabular data analysis.

A new framework improves graph representation using self-supervised learning.

New methods enhance detection and classification of software vulnerabilities.

Learn how neural networks analyze data structured as graphs.

A new framework for generating graphs using discrete representations and transformer models.

A new method improves modeling of potential energy surfaces in atomistic simulations.

A new framework enhances demand forecasting accuracy, tackling cold-start challenges effectively.

Exploring how location relationships change throughout the day for better planning.

A new approach enhances the comparison of proteins, aiding research and drug discovery.

A method to improve understanding of graph neural networks.

Explore the key concepts and approaches in dependency parsing for natural language processing.

A fresh approach to improving Graph Neural Networks using model soups.

A look into link prediction methods and their applications across various fields.

Assessing GSL methods for improved graph data learning.

A novel method to enhance graph representation and classification accuracy.

A new approach enhances GNNs by adapting to diverse datasets without retraining.

Introducing GMMD, a framework to enhance fairness in graph neural networks.

A new method enhances photon energy reconstruction in particle experiments.

New method creates large, detailed graphs for diverse applications.

A new tool speeds up graph generation for crystalline materials using modern technology.

A new method enhances multi-criteria recommendations using graph neural networks.

Geometric Pooling improves feature retention in graph data analysis.

A look at techniques to speed up Graph Neural Networks in complex datasets.

MepoGNN combines models to improve predictions of disease spread.

ViG-UNet combines graph neural networks and U-Net for improved medical image analysis.

A new method enhances cancer diagnosis using whole slide images and advanced learning techniques.

GIDS optimizes large-scale graph training for better efficiency and speed.

A new approach to improve multi-variate time-series predictions using GNNs.