{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/fuzzy-graph-neural-network-for-few-shot","title":"Fuzzy Graph Neural Network for Few-Shot Learning","arxiv_id":null,"date":"2020-07-19","proceeding":null,"authors":["Tong Wei","Junlin Hou and Rui Feng"],"abstract":"Recent works have shown that graph neural networks (GNNs) can substantially improve the performance of fewshot learning benefitting from their natural ability to learn interclass uniqueness and intra-class commonality. However, previous\r\nGNN methods have not achieved satisfactory performance due to\r\nthe absence of a strong relational inductive bias which determines\r\nhow entities interact and are isolated. In this paper, inspired by\r\nthe fuzzy theory, we propose a novel meta-learning method called\r\nFuzzy GNN (FGNN), which obtains superior relational inductive\r\nbiases in each episode, for few-shot learning. Specifically, we\r\nemploy an edge-focused GNN to perform the edge prediction\r\nby iteratively updating the edge-labels. According to the output\r\nof edge prediction, we design a fuzzy membership function\r\nto achieve more exact relationship representations for node\r\nclassification. The parameters of the FGNN are learned by\r\nepisodic training with mixed loss including node-label and edgelabel. Extensive experimental evaluation clearly demonstrates the\r\neffectiveness of FGNN. The results show that our method achieves\r\nstate-of-the-art performance and a significant improvement over\r\nother GNN methods on two few-shot learning benchmarks.","url_abs":"https://ieeexplore.ieee.org/document/9207213","url_pdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9207213","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"fuzzy-graph-neural-network-for-few-shot","repo_url":"https://github.com/sadbb/few-shot-fgnn","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"graph-neural-network","task_name":"Graph Neural Network"},{"task_slug":"inductive-bias","task_name":"Inductive Bias"},{"task_slug":"meta-learning","task_name":"Meta-Learning"},{"task_slug":"node-classification","task_name":"Node Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}