File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1137/1.9781611973440.24
- Scopus: eid_2-s2.0-84959888323
Supplementary
-
Citations:
- Scopus: 0
- Appears in Collections:
Conference Paper: Mixed-transfer: Transfer learning over mixed graphs
Title | Mixed-transfer: Transfer learning over mixed graphs |
---|---|
Authors | |
Keywords | Co-occurrence data Random walk Heterogeneous transfer learning |
Issue Date | 2014 |
Citation | SIAM International Conference on Data Mining 2014 (SDM 2014), Philadelphia, PA, 24-26 April 2014. In Proceedings of the 2014 SIAM International Conference on Data Mining, 2014, v. 1, p. 208-216 How to Cite? |
Abstract | Copyright © SIAM. Heterogeneous transfer learning has been proposed as a new learning strategy to improve performance in a target domain by leveraging data from other heterogeneous source domains where feature spaces can be different across different domains. In order to connect two different spaces, one common technique is to bridge feature spaces by using some co-occurrence data. For example, annotated images can be used to build feature mapping from words to image features, and then applied on text-to-image knowledge transfer. However, in practice, such co-occurrence data are often from Web, e.g. Flickr, and generated by users. That means these data can be sparse and contain personal biases. Directly building models based on them may fail to provide reliable bridge. To solve these aforementioned problems, in this paper, we propose a novel algorithm named Mixed-Transfer. It is composed of three components, that is, a cross domain harmonic function to avoid personal biases, a joint transition probability graph of mixed instances and features to model the heterogeneous transfer learning problem, a random walk process to simulate the label propagation on the graph and avoid the data sparsity problem. We conduct experiments on 171 real-world tasks, showing that the proposed approach outperforms four state-of-the-art heterogeneous transfer learning algorithms. |
Persistent Identifier | http://hdl.handle.net/10722/276718 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Tan, Ben | - |
dc.contributor.author | Zhong, Erheng | - |
dc.contributor.author | Ng, Michael K. | - |
dc.contributor.author | Yang, Qiang | - |
dc.date.accessioned | 2019-09-18T08:34:26Z | - |
dc.date.available | 2019-09-18T08:34:26Z | - |
dc.date.issued | 2014 | - |
dc.identifier.citation | SIAM International Conference on Data Mining 2014 (SDM 2014), Philadelphia, PA, 24-26 April 2014. In Proceedings of the 2014 SIAM International Conference on Data Mining, 2014, v. 1, p. 208-216 | - |
dc.identifier.uri | http://hdl.handle.net/10722/276718 | - |
dc.description.abstract | Copyright © SIAM. Heterogeneous transfer learning has been proposed as a new learning strategy to improve performance in a target domain by leveraging data from other heterogeneous source domains where feature spaces can be different across different domains. In order to connect two different spaces, one common technique is to bridge feature spaces by using some co-occurrence data. For example, annotated images can be used to build feature mapping from words to image features, and then applied on text-to-image knowledge transfer. However, in practice, such co-occurrence data are often from Web, e.g. Flickr, and generated by users. That means these data can be sparse and contain personal biases. Directly building models based on them may fail to provide reliable bridge. To solve these aforementioned problems, in this paper, we propose a novel algorithm named Mixed-Transfer. It is composed of three components, that is, a cross domain harmonic function to avoid personal biases, a joint transition probability graph of mixed instances and features to model the heterogeneous transfer learning problem, a random walk process to simulate the label propagation on the graph and avoid the data sparsity problem. We conduct experiments on 171 real-world tasks, showing that the proposed approach outperforms four state-of-the-art heterogeneous transfer learning algorithms. | - |
dc.language | eng | - |
dc.relation.ispartof | Proceedings of the 2014 SIAM International Conference on Data Mining | - |
dc.subject | Co-occurrence data | - |
dc.subject | Random walk | - |
dc.subject | Heterogeneous transfer learning | - |
dc.title | Mixed-transfer: Transfer learning over mixed graphs | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_OA_fulltext | - |
dc.identifier.doi | 10.1137/1.9781611973440.24 | - |
dc.identifier.scopus | eid_2-s2.0-84959888323 | - |
dc.identifier.volume | 1 | - |
dc.identifier.spage | 208 | - |
dc.identifier.epage | 216 | - |