Deep Partial Multi-View Learning

Deep Partial Multi-View Learning

Although multi-view learning has made significant progress over the past few decades, it is still challenging due to the difficulty in modeling complex correlations among different views, especially under the context of view missing. To address the challenge, we propose a novel framework termed Cros...

Ausführliche Beschreibung

Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on pattern analysis and machine intelligence. - 1979. - 44(2022), 5 vom: 12. Mai, Seite 2402-2415
1. Verfasser: Zhang, Changqing (VerfasserIn)
Weitere Verfasser: Cui, Yajie, Han, Zongbo, Zhou, Joey Tianyi, Fu, Huazhu, Hu, Qinghua
Format: Online-Aufsatz
Sprache:English
Veröffentlicht: 2022
Zugriff auf das übergeordnete Werk:IEEE transactions on pattern analysis and machine intelligence
Schlagworte:Journal Article
LEADER 01000naa a22002652 4500
001 NLM317482424
003 DE-627
005 20231225163348.0
007 cr uuu---uuuuu
008 231225s2022 xx |||||o 00| ||eng c
024 7 |a 10.1109/TPAMI.2020.3037734  |2 doi 
028 5 2 |a pubmed24n1058.xml 
035 |a (DE-627)NLM317482424 
035 |a (NLM)33180720 
040 |a DE-627  |b ger  |c DE-627  |e rakwb 
041 |a eng 
100 1 |a Zhang, Changqing  |e verfasserin  |4 aut 
245 1 0 |a Deep Partial Multi-View Learning 
264 1 |c 2022 
336 |a Text  |b txt  |2 rdacontent 
337 |a ƒaComputermedien  |b c  |2 rdamedia 
338 |a ƒa Online-Ressource  |b cr  |2 rdacarrier 
500 |a Date Revised 04.04.2022 
500 |a published: Print-Electronic 
500 |a Citation Status PubMed-not-MEDLINE 
520 |a Although multi-view learning has made significant progress over the past few decades, it is still challenging due to the difficulty in modeling complex correlations among different views, especially under the context of view missing. To address the challenge, we propose a novel framework termed Cross Partial Multi-View Networks (CPM-Nets), which aims to fully and flexibly take advantage of multiple partial views. We first provide a formal definition of completeness and versatility for multi-view representation and then theoretically prove the versatility of the learned latent representations. For completeness, the task of learning latent multi-view representation is specifically translated to a degradation process by mimicking data transmission, such that the optimal tradeoff between consistency and complementarity across different views can be implicitly achieved. Equipped with adversarial strategy, our model stably imputes missing views, encoding information from all views for each sample to be encoded into latent representation to further enhance the completeness. Furthermore, a nonparametric classification loss is introduced to produce structured representations and prevent overfitting, which endows the algorithm with promising generalization under view-missing cases. Extensive experimental results validate the effectiveness of our algorithm over existing state of the arts for classification, representation learning and data imputation 
650 4 |a Journal Article 
700 1 |a Cui, Yajie  |e verfasserin  |4 aut 
700 1 |a Han, Zongbo  |e verfasserin  |4 aut 
700 1 |a Zhou, Joey Tianyi  |e verfasserin  |4 aut 
700 1 |a Fu, Huazhu  |e verfasserin  |4 aut 
700 1 |a Hu, Qinghua  |e verfasserin  |4 aut 
773 0 8 |i Enthalten in  |t IEEE transactions on pattern analysis and machine intelligence  |d 1979  |g 44(2022), 5 vom: 12. Mai, Seite 2402-2415  |w (DE-627)NLM098212257  |x 1939-3539  |7 nnns 
773 1 8 |g volume:44  |g year:2022  |g number:5  |g day:12  |g month:05  |g pages:2402-2415 
856 4 0 |u http://dx.doi.org/10.1109/TPAMI.2020.3037734  |3 Volltext 
912 |a GBV_USEFLAG_A 
912 |a SYSFLAG_A 
912 |a GBV_NLM 
912 |a GBV_ILN_350 
951 |a AR 
952 |d 44  |j 2022  |e 5  |b 12  |c 05  |h 2402-2415