计算机与现代化

• 应用与开发 • 上一篇    下一篇

多标记学习研究综述

  

  1. 南京理工大学计算机科学与工程学院,江苏南京210094
  • 收稿日期:2015-03-09 出版日期:2015-08-08 发布日期:2015-08-19
  • 作者简介:梁伟超(1991-),男,江苏南京人,南京理工大学计算机科学与工程学院硕士研究生,研究方向:机器学习,数据挖掘; 宋斌(1968-),男,副教授,硕士,研究方向:数据挖掘,Web信息处理。

A Review on Multi-Label Learning

  1. School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
  • Received:2015-03-09 Online:2015-08-08 Published:2015-08-19

摘要: 多标记学习不同于传统的监督学习,它是为了解决客观世界中多义性对象的建模问题而提出的一种学习框架。在该框架下,一个示例可以同时隶属于多个标记。经过十多年的发展,机器学习界已经出现了大量关于多标记学习的研究成果,并得到了广泛的应用。本文对多标记学习问题进行系统而详细的阐述,给出多标记学习的问题定义和评价指标,重点介绍多标记学习算法,并提出多标记学习进一步的研究方向。

关键词: 多标记学习, 评价指标, 问题转换, 算法适应, 机器学习, 数据挖掘

Abstract: Multi-label learning is different from the traditional supervised learning. It is a framework which is proposed to represent objects which might have multiple semantic meanings simultaneously in the external world. Under this framework, an instance might be associated with a set of labels. Over the past decades, a lot of research results of multi-label learning have been achieved and gotten extensive application. This paper provides a systematic and detailed review in this area. Firstly, fundamentals on multi-label learning including formal definition and evaluation metrics are given. Secondly, some important and classic multi-label learning algorithms are presented. Finally, some valuable research directions in this area are discussed.

Key words: multi-label learning, evaluation metrics, problem transformation, algorithm adaptation, machine learning, data mining

中图分类号: