Graph Level Prediction

Graph-level prediction focuses on developing machine learning models that can accurately predict properties or labels of entire graphs, rather than individual nodes. Current research explores various approaches, including those based on random walks, the analysis of attributed graphlets (small subgraphs with associated features), and the application of graph neural networks (GNNs). These methods aim to improve prediction accuracy and interpretability, addressing challenges such as over-smoothing in GNNs and the computational cost of exploring all possible subgraphs. The ability to effectively predict graph-level properties has significant implications for diverse fields, including drug discovery (molecular graphs), autonomous driving (traffic networks), and social network analysis.

Papers

July 1, 2024

Revisiting Random Walks for Learning on Graphs
Jinwoo Kim, Olga Zaghen, Ayhan Suleymanzade, Youngmin Ryou, Seunghoon Hong
Graph Drawing Markov Chain Random Walk Graph Level Prediction

February 10, 2024

Learning Attributed Graphlets: Predictive Graph Mining by Graphlets with Trainable Attribute
Tajima Shinji, Ren Sugihara, Ryota Kitahara, Masayuki Karasuyama
Interpretable Model Tree Search Interpretable Classifier Graph Level Prediction

January 8, 2024

Predicting the structure of dynamic graphs
Sevvandi Kandanaarachchi
Graph Drawing Link Prediction Dynamic Graph Inner Structure Graph Level Prediction

December 10, 2023

Graph-based Prediction and Planning Policy Network (GP3Net) for scalable self-driving in dynamic environments using Deep Reinforcement Learning
Jayabrata Chowdhury, Venkataramanan Shivaraman, Suresh Sundaram, P B Sujit
Deep Reinforcement Learning Autonomous Vehicle Learning Based Dynamic Environment Non Stationary Environment Policy Network Expert Driver Graph Level Prediction Scalable Autonomous Driving

January 26, 2023

WL meet VC
Christopher Morris, Floris Geerts, Jan Tönshoff, Martin Grohe
Graph Neural Network Goal Vertex Weisfeiler Leman Vapnik Chervonenkis Dimension Graph Level Prediction

March 7, 2022

Graph Neural Networks for Image Classification and Reinforcement Learning using Graph representations
Naman Goyal, David Steiner
Reinforcement Learning Graph Neural Network Computer Vision Image Classification Graph Representation Graph Level Prediction