Saved in:
Bibliographic Details
Main Authors: Yuan, Qiao, Guan, Sheng-Uei, Ni, Pin, Luo, Tianlun, Man, Ka Lok, Wong, Prudence, Chang, Victor
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2301.12230
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911549061332992
author Yuan, Qiao
Guan, Sheng-Uei
Ni, Pin
Luo, Tianlun
Man, Ka Lok
Wong, Prudence
Chang, Victor
author_facet Yuan, Qiao
Guan, Sheng-Uei
Ni, Pin
Luo, Tianlun
Man, Ka Lok
Wong, Prudence
Chang, Victor
contents Continual Graph Learning (CGL) enables models to incrementally learn from streaming graph-structured data without forgetting previously acquired knowledge. Experience replay is a common solution that reuses a subset of past samples during training. However, it may lead to information loss and privacy risks. Generative replay addresses these concerns by synthesizing informative subgraphs for rehearsal. Existing generative replay approaches often rely on graph condensation via distribution matching, which faces two key challenges: (1) the use of random feature encodings may fail to capture the characteristic kernel of the discrepancy metric, weakening distribution alignment; and (2) matching over a fixed small subgraph cannot guarantee low risk on previous tasks, as indicated by domain adaptation theory. To overcome these limitations, we propose an Adversarial Condensation based Generative Replay (ACGR) framwork. It reformulates graph condensation as a min-max optimization problem to achieve better distribution matching. Moreover, instead of learning a single subgraph, we learn its distribution, allowing for the generation of multiple samples and improved empirical risk minimization. Experiments on three benchmark datasets demonstrate that ACGR outperforms existing methods in both accuracy and stability.
format Preprint
id arxiv_https___arxiv_org_abs_2301_12230
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Continual Graph Learning: A Survey
Yuan, Qiao
Guan, Sheng-Uei
Ni, Pin
Luo, Tianlun
Man, Ka Lok
Wong, Prudence
Chang, Victor
Machine Learning
Artificial Intelligence
Continual Graph Learning (CGL) enables models to incrementally learn from streaming graph-structured data without forgetting previously acquired knowledge. Experience replay is a common solution that reuses a subset of past samples during training. However, it may lead to information loss and privacy risks. Generative replay addresses these concerns by synthesizing informative subgraphs for rehearsal. Existing generative replay approaches often rely on graph condensation via distribution matching, which faces two key challenges: (1) the use of random feature encodings may fail to capture the characteristic kernel of the discrepancy metric, weakening distribution alignment; and (2) matching over a fixed small subgraph cannot guarantee low risk on previous tasks, as indicated by domain adaptation theory. To overcome these limitations, we propose an Adversarial Condensation based Generative Replay (ACGR) framwork. It reformulates graph condensation as a min-max optimization problem to achieve better distribution matching. Moreover, instead of learning a single subgraph, we learn its distribution, allowing for the generation of multiple samples and improved empirical risk minimization. Experiments on three benchmark datasets demonstrate that ACGR outperforms existing methods in both accuracy and stability.
title Continual Graph Learning: A Survey
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2301.12230