计算机科学
极限学习机
机器学习
人工智能
水准点(测量)
支持向量机
数据流
半监督学习
多标签分类
核(代数)
标记数据
数据流挖掘
监督学习
数据挖掘
模式识别(心理学)
人工神经网络
数学
地理
组合数学
电信
大地测量学
作者
Shiyuan Qiu,Peipei Li,Xuegang Hu
标识
DOI:10.1109/ijcnn55064.2022.9892701
摘要
Semi-supervised multi-label data stream classification serves a practical yet challenging task since only a small number of labeled instances are available in real streaming environments. However, the mainstream of existing semi-supervised multi-label classification technique is focused on the batch process. Meanwhile, many data stream classification approaches have been proposed, and one of popularly used base models is ELM (extreme learning machine). But only few ELM-based algorithms are proposed for the multi-label data stream classification. Therefore, in this paper, we present a novel Semi-supervised Online Extreme Learning Machine with Kernel function for multi-label data stream classification, called SSO-KELM. Specifically, we firstly introduce the kernel function to output the multi-dimensional vector, for adapting to the multi-label data. Secondly, to make full use of labeled and unlabeled data in a data stream, we derive a novel online semi-supervised ELM algorithm, which can adapt to the stream setting and achieve a higher classification accuracy. Finally, extensive experiments conducted on six benchmark multi-label data sets demonstrate the effectiveness of our approach compared to state-of-the-art approaches.
科研通智能强力驱动
Strongly Powered by AbleSci AI