Cross-domain metric learning based on information theory
文献类型：会议论文
| 作者 | Wang, Hao (1) ; Wang, Wei (2) ; Zhang, Chen (2) ; Xu, Fanjiang (2) |
| 出版日期 | 2014 |
| 会议名称 | 28th AAAI Conference on Artificial Intelligence, AAAI 2014, 26th Innovative Applications of Artificial Intelligence Conference, IAAI 2014 and the 5th Symposium on Educational Advances in Artificial Intelligence, EAAI 2014 |
| 会议日期 | July 27, 2014 - July 31, 2014 |
| 会议地点 | Quebec City, QC, Canada |
| 页码 | 2099-2105 |
| 通讯作者 | Wang, Hao |
| 中文摘要 | Supervised metric learning plays a substantial role in statistical classification. Conventional metric learning algorithms have limited utility when the training data and testing data are drawn from related but different domains (i.e., source domain and target domain). Although this issue has got some progress in feature-based transfer learning, most of the work in this area suffers from non-trivial optimization and pays little attention to preserving the discriminating information. In this paper, we propose a novel metric learning algorithm to transfer knowledge from the source domain to the target domain in an information-theoretic setting, where a shared Mahalanobis distance across two domains is learnt by combining three goals together: 1) reducing the distribution difference between different domains; 2) preserving the geometry of target domain data; 3) aligning the geometry of source domain data with its label information. Based on this combination, the learnt Mahalanobis distance effectively transfers the discriminating power and propagates standard classifiers across these two domains. More importantly, our proposed method has closed-form solution and can be efficiently optimized. Experiments in two real-world applications demonstrate the effectiveness of our proposed method. |
| 英文摘要 | Supervised metric learning plays a substantial role in statistical classification. Conventional metric learning algorithms have limited utility when the training data and testing data are drawn from related but different domains (i.e., source domain and target domain). Although this issue has got some progress in feature-based transfer learning, most of the work in this area suffers from non-trivial optimization and pays little attention to preserving the discriminating information. In this paper, we propose a novel metric learning algorithm to transfer knowledge from the source domain to the target domain in an information-theoretic setting, where a shared Mahalanobis distance across two domains is learnt by combining three goals together: 1) reducing the distribution difference between different domains; 2) preserving the geometry of target domain data; 3) aligning the geometry of source domain data with its label information. Based on this combination, the learnt Mahalanobis distance effectively transfers the discriminating power and propagates standard classifiers across these two domains. More importantly, our proposed method has closed-form solution and can be efficiently optimized. Experiments in two real-world applications demonstrate the effectiveness of our proposed method. |
| 收录类别 | EI |
| 会议录出版地 | AI Access Foundation |
| 语种 | 英语 |
| ISBN号 | 9781577356790 |
| 源URL | [http://ir.iscas.ac.cn/handle/311060/16594]  |
| 专题 | 软件研究所_软件所图书馆_会议论文 |
| 推荐引用方式 GB/T 7714 | Wang, Hao ,Wang, Wei ,Zhang, Chen ,et al. Cross-domain metric learning based on information theory[C]. 见:28th AAAI Conference on Artificial Intelligence, AAAI 2014, 26th Innovative Applications of Artificial Intelligence Conference, IAAI 2014 and the 5th Symposium on Educational Advances in Artificial Intelligence, EAAI 2014. Quebec City, QC, Canada. July 27, 2014 - July 31, 2014. |
入库方式： OAI收割
来源：软件研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。