Summary of Inductive Graph Few-shot Class Incremental Learning, by Yayong Li et al.
Inductive Graph Few-shot Class Incremental Learning
by Yayong Li, Peyman Moghadam, Can Peng, Nan Ye, Piotr Koniusz
First submitted to arxiv on: 11 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a novel approach to graph-based incremental learning, focusing on Graph Few-Shot Class Incremental Learning (GFSCIL). The authors introduce inductive GFSCIL, which learns to classify new classes without accessing previous data, addressing the practical concern of transductive methods that require storing historical data. To tackle catastrophic forgetting and overfitting issues, they propose Topology-based class Augmentation and Prototype calibration (TAP), a method that combines multi-topology class augmentation and iterative prototype calibration to improve model generalization and adapt to changing feature distributions. The authors demonstrate their approach on four datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about teaching computers to learn new things as more information becomes available, without forgetting what they already know. They’re trying to solve a problem where the computer has to keep learning and adapting as new data comes in, but it’s hard because the old data isn’t available anymore. To fix this, they created a new way of doing things that uses three different steps: making the model more versatile, improving how well it can tell classes apart, and helping the old classes adapt to the changing information. They tested their idea on four datasets. |
Keywords
» Artificial intelligence » Few shot » Generalization » Overfitting
