EUVIMEDEUVIMED entdecken
Crossref

Hierarchical Prototype Networks for Continual Graph Representation Learning

Xikun Zhang, Dongjin Song, Dacheng Tao

IEEE Transactions on Pattern Analysis and Machine Intelligence · 01.04.2023 · 45 (4) · 4622-4636

DOI: 10.1109/tpami.2022.3186909 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, erschienen am 01.04.2023. Verfasst von Xikun Zhang und 2 weiteren Autor:innen. Die Arbeit wurde bisher 56-mal zitiert (Crossref) und nennt 62 Literaturangaben. Gefördert von: Australian Research Council Projects, UConn Scholarship Facilitation Fund, National Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Despite significant advances in graph representation learning, little attention has been paid to the more practical continual learning scenario in which new categories of nodes (e.g., new research areas in citation networks, or new types of products in co-purchasing networks) and their associated edges are continuously emerging, causing catastrophic forgetting on previous categories. Existing methods either ignore the rich topological information or sacrifice plasticity for stability. To this end, we present Hierarchical Prototype Networks (HPNs) which extract different levels of abstract knowledge in the form of prototypes to represent the continuously expanded graphs. Specifically, we first leverage a set of Atomic Feature Extractors (AFEs) to encode both the elemental attribute information and the topological structure of the target node. Next, we develop HPNs to adaptively select relevant AFEs and represent each node with three levels of prototypes. In this way, whenever a new category of nodes is given, only the relevant AFEs and prototypes at each level will be activated and refined, while others remain uninterrupted to maintain the performance over existing nodes. Theoretically, we first demonstrate that the memory consumption of HPNs is bounded regardless of how many tasks are encountered. Then, we prove that under mild constraints, learning new tasks will not alter the prototypes matched to previous data, thereby eliminating the forgetting problem. The theoretical results are supported by experiments on five datasets, showing that HPNs not only outperform state-of-the-art baseline techniques but also consume relatively less memory. Code and datasets are available at https://github.com/QueuQ/HPNs.

Förderung

  • Australian Research Council Projects
  • UConn Scholarship Facilitation Fund
  • National Science Foundation
  • National Institute of Food and Agriculture

Zitieren

APA

Zhang, X., Song, D., & Tao, D. (2023). Hierarchical Prototype Networks for Continual Graph Representation Learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(4), 4622-4636. https://doi.org/10.1109/tpami.2022.3186909

Vancouver

Zhang X, Song D, Tao D. Hierarchical Prototype Networks for Continual Graph Representation Learning. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2023;45(4):4622-4636. doi:10.1109/tpami.2022.3186909

BibTeX
@article{Zhang2023,
  author = {Zhang, Xikun and Song, Dongjin and Tao, Dacheng},
  title = {Hierarchical Prototype Networks for Continual Graph Representation Learning},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2023},
  volume = {45},
  number = {4},
  pages = {4622--4636},
  doi = {10.1109/tpami.2022.3186909}
}
Kostenlos starten →