Graph Based Active Learning

Graph-based active learning aims to efficiently train machine learning models on graph-structured data by strategically selecting a minimal subset of nodes for labeling. Current research focuses on developing novel acquisition functions that consider both node informativeness (e.g., uncertainty, diversity) and representativeness within the graph structure, often incorporating graph neural networks (GNNs) and graph signal processing techniques. This approach is crucial for applications where labeling data is expensive or time-consuming, improving the efficiency and effectiveness of various tasks, including node classification, entity cluster repair, and even antimicrobial peptide design. The resulting label-efficient models offer significant advantages across diverse domains.

Papers

October 24, 2024

Perturbation-based Graph Active Learning for Weakly-Supervised Belief Representation Learning
Dachun Sun, Ruijie Wang, Jinning Li, Ruipeng Han, Xinyi Liu, You Lyu, Tarek Abdelzaher
Representation Learning Weakly Supervised Social Graph Graph Perturbation Graph Based Active Learning Semi Supervision

September 14, 2024

MALADY: Multiclass Active Learning with Auction Dynamics on Graphs
Gokul Bhusal, Kevin Miller, Ekaterina Merkurjev
Graph Drawing Active Learning Semi Supervised Learning Efficient Active Learning Graph Based Active Learning Multiclass Learning

August 15, 2024

Robust Offline Active Learning on Graphs
Yuanchen Wu, Yubai Yuan
Graph Drawing Active Learning Graph Signal Graph Based Active Learning

January 26, 2024

Graph-based Active Learning for Entity Cluster Repair
Victor Christen, Daniel Obraczka, Marvin Hofer, Martin Franke, Erhard Rahm
Data Deduplication Graph Based Active Learning Entity Cluster

August 17, 2023

Mitigating Semantic Confusion from Hostile Neighborhood for Graph Active Learning
Tianmeng Yang, Min Zhou, Yujing Wang, Zhengjie Lin, Lujia Pan, Bin Cui, Yunhai Tong
Graph Neural Network Node Classification Semantic Feature Active Learning Framework Graph Based Active Learning

July 31, 2023

DiffusAL: Coupling Active Learning with Graph Diffusion for Label-Efficient Node Classification
Sandra Gilhuber, Julian Busch, Daniel Rotthues, Christian M. M. Frey, Thomas Seidl
Active Learning Node Classification Graph Diffusion Graph Based Active Learning Active Directed Graph

December 5, 2022

Dissimilar Nodes Improve Graph Active Learning
Zhicheng Ren, Yifu Yuan, Yuxin Wu, Xiaxuan Gao, Yewen Wang, Yizhou Sun
Node Similarity Dissimilarity Measure Unlabeled Node Structural Similarity Index Graph Based Active Learning Active Directed Graph

December 2, 2022

SMARTQUERY: An Active Learning Framework for Graph Neural Networks through Hybrid Uncertainty Reduction
Xiaoting Li, Yuhang Wu, Vineeth Rakesh, Yusan Lin, Hao Yang, Fei Wang
Graph Neural Network Active Learning Uncertainty Estimation Graph Structure Label Propagation Active Learning Framework Graph Based Active Learning

October 27, 2022

Poisson Reweighted Laplacian Uncertainty Sampling for Graph-based Active Learning
Kevin Miller, Jeff Calder
Active Learning Uncertainty Sampling Graph Based Active Learning

September 18, 2022

Graph-Based Active Machine Learning Method for Diverse and Novel Antimicrobial Peptides Generation and Selection
Bonaventure F. P. Dossou, Dianbo Liu, Xu Ji, Moksh Jain, Almer M. van der Sloot, Roger Palou, Michael Tyers, Yoshua Bengio
View Selection Diverse Set Graph Based Active Learning Antibiotic Resistance Large Scale Screening Antimicrobial Peptide

March 31, 2022

Graph-based Active Learning for Semi-supervised Classification of SAR Data
Kevin Miller, John Mauro, Jason Setiadi, Xoaquin Baca, Zhan Shi, Jeff Calder, Andrea L. Bertozzi
Semi Supervised Synthetic Aperture Radar Synthetic Aperture Radar Data Graph Based Learning Semi Supervised Classification Graph Based Active Learning

March 2, 2022

Information Gain Propagation: a new way to Graph Active Learning with Soft Labels
Wentao Zhang, Yexin Wang, Zhenbang You, Meng Cao, Ping Huang, Jiulong Shan, Zhi Yang, Bin Cui
Graph Neural Network Active Learning Soft Label New Way Information Gain Label Efficiency Graph Based Active Learning

February 3, 2022

GALAXY: Graph-based Active Learning at the Extreme
Jifan Zhang, Julian Katz-Samuels, Robert Nowak
Active Learning Distant GALAXY Balanced Dataset Extreme Value Graph Based Active Learning Class Balancing

Graph Based Active Learning

Papers

Perturbation-based Graph Active Learning for Weakly-Supervised Belief Representation Learning

MALADY: Multiclass Active Learning with Auction Dynamics on Graphs

Robust Offline Active Learning on Graphs

Graph-based Active Learning for Entity Cluster Repair

Mitigating Semantic Confusion from Hostile Neighborhood for Graph Active Learning

DiffusAL: Coupling Active Learning with Graph Diffusion for Label-Efficient Node Classification

Dissimilar Nodes Improve Graph Active Learning

SMARTQUERY: An Active Learning Framework for Graph Neural Networks through Hybrid Uncertainty Reduction

Poisson Reweighted Laplacian Uncertainty Sampling for Graph-based Active Learning

Graph-Based Active Machine Learning Method for Diverse and Novel Antimicrobial Peptides Generation and Selection

Graph-based Active Learning for Semi-supervised Classification of SAR Data

Information Gain Propagation: a new way to Graph Active Learning with Soft Labels

GALAXY: Graph-based Active Learning at the Extreme