Graphsage
PulseAugur coverage of Graphsage — every cluster mentioning Graphsage across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New study compares AI explanation methods for relational databases
Researchers have developed a new methodology to compare different importance attribution methods for deep learning models operating on relational databases. Their study analyzed how these methods rank and score attribut…
-
New DPR method slashes GNN training memory, boosts inference speed
Researchers have developed a novel method called Decoupled Prune-and-Reconstruct (DPR) to address memory limitations in training large graph neural networks (GNNs). DPR employs a double-layer sparsification technique wh…
-
BehaveRec framework improves recommendation systems by analyzing long-term user behavior · 2 sources tracked
Researchers have developed BehaveRec, a new framework for sequential recommendation systems that analyzes user behavior over a 90-day window and incorporates community signals. Unlike previous models that assumed long-t…
-
New GraphBEAT Framework Enhances Cancer Risk Stratification from EHRs
Researchers have developed GraphBEAT, a novel temporal graph representation learning framework designed to improve early risk stratification for gastrointestinal cancers using electronic health records (EHRs). This meth…
-
New AneumoBench dataset evaluates synthetic-geometry transfer for aneurysm CFD
Researchers have introduced AneumoBench, a new benchmark dataset designed to evaluate synthetic-geometry transfer for aneurysm computational fluid dynamics (CFD). The dataset links 401 source aneurysm geometries to over…
-
New PEBDSAM Module Enhances Graph Neural Networks
Researchers have introduced a novel Position Encoding-Based Deformable Spatial Aggregation Module (PEBDSAM) designed to enhance Graph Neural Networks (GNNs). This module addresses fundamental limitations in traditional …
-
Reification method enables zero-shot link prediction for GNNs
Researchers have developed a novel method called "reification" to enable graph neural networks (GNNs) to perform zero-shot link prediction on unseen graphs. This technique transforms graph data into a fixed vocabulary o…
-
Graph Neural Networks Improve Road Network Disruption Analysis
Researchers have developed graph neural networks (GNNs) to efficiently estimate connectivity loss in road networks following disruptions. The study compares GCN, GraphSAGE, and MPNN models, finding that residual GCN and…
-
GENIE watermarking scheme protects GNNs for link prediction
Researchers have developed GENIE, a novel watermarking scheme designed to protect Graph Neural Network (GNN) models used for link prediction. Unlike previous methods that focused on node or graph classification, GENIE a…
-
GNNs suffer temporal leakage in financial models; new benchmark released
Researchers have identified widespread temporal leakage in message-passing graph neural networks (GNNs) when applied to financial transaction data. This leakage occurs because standard training splits can expose the mod…
-
Pinterest Ads leverage graph embeddings for improved CTR and CVR
Researchers have developed a novel approach to enhance advertising models by integrating user onsite and offsite conversion data into a large-scale heterogeneous graph. This method utilizes a Knowledge Graph Embedding (…
-
Network analysis improves unreliable news detection
Researchers have developed a new method for detecting unreliable news domains by analyzing URL-sharing patterns on Telegram. They constructed a domain co-sharing network, revealing that unreliable and reliable news doma…
-
New protocol for GNN cross-task transfer reveals directional predictability
Researchers have developed a new protocol to reliably evaluate cross-task transfer in Graph Neural Networks (GNNs) for node classification (NC) and link prediction (LP) tasks. Their findings indicate that transfer from …
-
New SKGFusionKAN method enhances IoT network intrusion detection using GNNs and KAN
Researchers have developed a new approach called SKGFusionKAN to improve intrusion detection in Internet of Things (IoT) networks. This method combines graph neural networks (GNNs), specifically GraphSage, with Kolmogor…
-
AI system AgentsCAD automates 3D print design modifications
Researchers have developed AgentsCAD, a multi-agent system that uses Large Language Models (LLMs) to automate design modifications for 3D-printed parts. The system analyzes STEP files, identifies potential manufacturing…
-
New AgentsCAD system automates FDM part design with LLM reasoning
Researchers have developed AgentsCAD, a novel multi-agent system designed to automate Design for Additive Manufacturing (DFAM) modifications for Fused Deposition Modeling (FDM) parts. The system processes STEP files, id…
-
Node embedding dimensionality impacts stability, new arXiv paper finds
A new arXiv paper explores how the dimensionality of node embeddings impacts their stability across different training runs and methods. Researchers found that dimensionality can significantly affect stability, with eff…
-
Deep-learning tool rapidly assesses post-hurricane grid damage and schedules repairs
Researchers have developed a two-stage deep-learning tool to expedite post-hurricane damage assessment and repair scheduling for electrical grids. The first stage identifies damaged lines using models like MLP, ResMLP, …
-
New GNN approach enhances multi-site pollution prediction accuracy
Researchers have developed a novel approach using Graph Neural Networks (GNNs) to improve the accuracy of particulate matter (PM) pollution prediction. This method dynamically constructs graphs based on inter-class rela…
-
Gradient leakage attacks threaten GNNs in circuit design
A new research paper details the first comprehensive evaluation of gradient leakage attacks (GLAs) on graph neural networks (GNNs) used in circuit design and hardware security. The study reveals that GLAs can expose sen…