|
|||||
|
|
| 面向共享数据的迁移组概率学习机 | |
| 引用本文： | 倪彤光 王士同 史荧中 张景祥. 面向共享数据的迁移组概率学习机[J]. 控制与决策, 2014, 29(8): 1363-1371 |
| 作者姓名： | 倪彤光 王士同 史荧中 张景祥 |
| 作者单位： | 1. 江南大学数字媒体学院，江苏无锡214122； 2. 常州大学信息科学与工程学院，江苏常州213164. |
| 基金项目： | 国家自然科学基金项目(61272210, 61170122)；江苏省自然科学基金项目(BK2012552). |
| 摘 要： | 为了解决机器学习中的主观信息缺失问题,提出一种新的面向共享数据的迁移组概率学习机(TGPLM-CD).该方法基于结构风险最小化模型,将源领域所含知识和目标领域的类标签组概率信息,特别是领域间的共享数据纳入学习框架中,实现了源领域和目标领域的知识迁移,在待研究领域数据信息不足的情况下提高了分类精确度.大量数据集上的实验结果验证了所提出方法的有效性. |
| 关 键 词： | 迁移学习 分类 支持向量机 共享数据 |
| 收稿时间： | 2013-04-01 |
| 修稿时间： | 2013-06-27 |
Transfer learning machine based on group probabilities toward common data |
|
| NI Tong-guang WANG Shi-tong SHI Ying-zhong ZHANG Jing-xiang. Transfer learning machine based on group probabilities toward common data[J]. Control and Decision, 2014, 29(8): 1363-1371 | |
| Authors: | NI Tong-guang WANG Shi-tong SHI Ying-zhong ZHANG Jing-xiang |
| Abstract: | To address the problem of man-made information scarcity in the machine learning, a novel transfer learning machine based on group probabilities toward the common data, called TGPLM-CD, is proposed. The proposed method is based on the structure risk minimization model, and both knowledges of the source domain and class label group probabilities are considered as well as the common data between the source domain and the target domain in the learning process, which realizes knowledge transfer between the source domain and the target domain. Experiment results on extensive datasets show the effectiveness of the proposed method. |
| Keywords: | Common Data |
| 本文献已被 CNKI 等数据库收录！ | |
| 点击此处可从《控制与决策》浏览原始摘要信息 | |
| 点击此处可从《控制与决策》下载全文 | |