Node Feature

Node features, descriptive attributes associated with individual nodes in a graph, are crucial for effective graph-based learning, yet often incomplete or noisy in real-world datasets. Current research focuses on developing robust methods for handling missing or imperfect node features, including techniques like feature imputation using graph structure and node embeddings, and the design of graph neural networks (GNNs) that incorporate multiple feature types and are resilient to noise. These advancements are significant because accurate and complete node features are essential for improving the performance of GNNs across diverse applications, ranging from link prediction and node classification to complex networked system modeling.

Papers

January 23, 2023

Explaining Deep Learning Hidden Neuron Activations using Concept Induction
Abhilekha Dalal, Md Kamruzzaman Sarker, Adrita Barua, Pascal Hitzler
Explainable AI Concept Discovery Node Feature Hidden Neuron Hidden Neuron Activation

January 9, 2023

SCENE: Reasoning about Traffic Scenes using Heterogeneous Graph Neural Networks
Thomas Monninger, Julian Schmidt, Jan Rupprecht, David Raba, Julian Jordan, Daniel Frank, Steffen Staab, Klaus Dietmayer
Node Classification Heterogeneous Graph Graph Convolution Heterogeneous Graph Neural Network Scene Structure Traffic Scene Node Feature

June 21, 2022

Nimble GNN Embedding with Tensor-Train Decomposition
Chunxing Yin, Da Zheng, Israt Nisa, Christos Faloutos, George Karypis, Richard Vuduc
Graph Neural Network Jina Embeddings GNN Based Node Feature Graph Partitioning

June 16, 2022

A Robust Stacking Framework for Training Deep Graph Models with Multifaceted Node Features
Jiuhai Chen, Jonas Mueller, Vassilis N. Ioannidis, Tom Goldstein, David Wipf
Graph Neural Network Layer Wise Node Feature Graph Datasets Graph Propagation Stacking Framework Deep Graph Model

June 1, 2022

Neural Decoding with Optimization of Node Activations
Eliya Nachmani, Yair Be'ery
Optimization Purpose Belief Propagation Error Correction Node Feature Neural Network Decoder Neural Decoding Teacher Decoder Novel Loss

May 18, 2022

Maslow's Hammer for Catastrophic Forgetting: Node Re-Use vs Node Activation
Sebastian Lee, Stefano Sarao Mannelli, Claudia Clopath, Sebastian Goldt, Andrew Saxe
Neural Network Continual LEArning Catastrophic Forgetting Anti Forgetting Node Feature Self Similar Feature Reuse

December 3, 2021

Structure-Aware Multi-Hop Graph Convolution for Graph Neural Networks
Yang Li, Yuichi Tanaka
Graph Neural Network Graph Convolution Node Feature Relational Embeddings Feature Distance Hop Neighbor

November 23, 2021

On the Unreasonable Effectiveness of Feature propagation in Learning on Graphs with Missing Node Features
Emanuele Rossi, Henry Kenlay, Maria I. Gorinova, Benjamin Paul Chamberlain, Xiaowen Dong, Michael Bronstein
LeArning Abstract Graph Neural Network Graph Drawing Node Classification Graph Machine Learning Feature Propagation Node Feature