File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/MASS.2019.00040
- Scopus: eid_2-s2.0-85085016773
Supplementary
-
Citations:
- Scopus: 0
- Appears in Collections:
Conference Paper: Deep neural network ensembles against deception: Ensemble diversity, accuracy and robustness
Title | Deep neural network ensembles against deception: Ensemble diversity, accuracy and robustness |
---|---|
Authors | |
Issue Date | 2019 |
Citation | Proceedings - 2019 IEEE 16th International Conference on Mobile Ad Hoc and Smart Systems, MASS 2019, 2019, p. 274-282 How to Cite? |
Abstract | Ensemble learning is a methodology that integrates multiple DNN learners for improving prediction performance of individual learners. Diversity is greater when the errors of the ensemble prediction is more uniformly distributed. Greater diversity is highly correlated with the increase in ensemble accuracy. Another attractive property of diversity optimized ensemble learning is its robustness against deception: an adversarial perturbation attack can mislead one DNN model to misclassify but may not fool other ensemble DNN members consistently. In this paper we first give an overview of the concept of ensemble diversity and examine the three types of ensemble diversity in the context of DNN classifiers. We then describe a set of ensemble diversity measures, a suite of algorithms for creating diversity ensembles and for performing ensemble consensus (voted or learned) for generating high accuracy ensemble output by strategically combining outputs of individual members. This paper concludes with a discussion on a set of open issues in quantifying ensemble diversity for robust deep learning. |
Persistent Identifier | http://hdl.handle.net/10722/343300 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Liu, Ling | - |
dc.contributor.author | Wei, Wenqi | - |
dc.contributor.author | Chow, Ka Ho | - |
dc.contributor.author | Loper, Margaret | - |
dc.contributor.author | Gursoy, Emre | - |
dc.contributor.author | Truex, Stacey | - |
dc.contributor.author | Wu, Yanzhao | - |
dc.date.accessioned | 2024-05-10T09:07:01Z | - |
dc.date.available | 2024-05-10T09:07:01Z | - |
dc.date.issued | 2019 | - |
dc.identifier.citation | Proceedings - 2019 IEEE 16th International Conference on Mobile Ad Hoc and Smart Systems, MASS 2019, 2019, p. 274-282 | - |
dc.identifier.uri | http://hdl.handle.net/10722/343300 | - |
dc.description.abstract | Ensemble learning is a methodology that integrates multiple DNN learners for improving prediction performance of individual learners. Diversity is greater when the errors of the ensemble prediction is more uniformly distributed. Greater diversity is highly correlated with the increase in ensemble accuracy. Another attractive property of diversity optimized ensemble learning is its robustness against deception: an adversarial perturbation attack can mislead one DNN model to misclassify but may not fool other ensemble DNN members consistently. In this paper we first give an overview of the concept of ensemble diversity and examine the three types of ensemble diversity in the context of DNN classifiers. We then describe a set of ensemble diversity measures, a suite of algorithms for creating diversity ensembles and for performing ensemble consensus (voted or learned) for generating high accuracy ensemble output by strategically combining outputs of individual members. This paper concludes with a discussion on a set of open issues in quantifying ensemble diversity for robust deep learning. | - |
dc.language | eng | - |
dc.relation.ispartof | Proceedings - 2019 IEEE 16th International Conference on Mobile Ad Hoc and Smart Systems, MASS 2019 | - |
dc.title | Deep neural network ensembles against deception: Ensemble diversity, accuracy and robustness | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/MASS.2019.00040 | - |
dc.identifier.scopus | eid_2-s2.0-85085016773 | - |
dc.identifier.spage | 274 | - |
dc.identifier.epage | 282 | - |