A robust self-training algorithm based on relative node graph

doi:10.1007/s10489-024-06062-0

	A robust self-training algorithm based on relative node graph
	Wang, Jikui1 ; Duan, Huiyu1 ; Zhang, Cuihong1 ; Nie, Feiping2
	2025-01
发表期刊	APPLIED INTELLIGENCE
卷号	55 期号:1
摘要	Self-training algorithm is a well-known framework of semi-supervised learning. How to select high-confidence samples is the key step for self-training algorithm. If high-confidence examples with incorrect labels are employed to train the classifier, the error will get worse during iterations. To improve the quality of high-confidence samples, a novel data editing technique termed Relative Node Graph Editing (RNGE) is put forward. Say concretely, mass estimation is used to calculate the density and peak of each sample to build a prototype tree to reveal the underlying spatial structure of the data. Then, we define the Relative Node Graph (RNG) for each sample. Finally, the mislabeled samples in the candidate high-confidence sample set are identified by hypothesis test based on RNG. Combined above, we propose a Robust Self-training Algorithm based on Relative Node Graph (STRNG), which uses RNGE to identify mislabeled samples and edit them. The experimental results show that the proposed algorithm can improve the performance of the self-training algorithm.
关键词	Semi-supervised learning High-confidence samples Self-training Data editing
DOI	10.1007/s10489-024-06062-0
收录类别	SCIE ; EI
ISSN	0924-669X
语种	英语
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:001355646600001
出版者	SPRINGER
EI入藏号	20244717399144
EI主题词	Self-supervised learning
EI分类号	1101.2 ; 1101.2.1 ; 1103.3 ; 1105.3 ; 1201.8
原始文献类型	Article
EISSN	1573-7497
引用统计	被引频次[WOS]：0 [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.lzufe.edu.cn/handle/39EH0E1M/38334
专题	信息工程与人工智能学院
通讯作者	Wang, Jikui
作者单位	1.Lanzhou Univ Finance & Econ, Coll Informat Engn & Artificial Intelligence, Lanzhou 730020, Gansu, Peoples R China; 2.Northwestern Polytech Univ, Sch Artificial Intelligence OPt & Elect IOPEN, Xian 710072, Shanxi, Peoples R China
第一作者单位	兰州财经大学
通讯作者单位	兰州财经大学
推荐引用方式 GB/T 7714	Wang, Jikui,Duan, Huiyu,Zhang, Cuihong,et al. A robust self-training algorithm based on relative node graph[J]. APPLIED INTELLIGENCE,2025,55(1).
APA	Wang, Jikui,Duan, Huiyu,Zhang, Cuihong,&Nie, Feiping.(2025).A robust self-training algorithm based on relative node graph.APPLIED INTELLIGENCE,55(1).
MLA	Wang, Jikui,et al."A robust self-training algorithm based on relative node graph".APPLIED INTELLIGENCE 55.1(2025).