Zobrazeno 1 - 10 of 21 pro vyhledávání: '"Saito, Sean"'
2
Report
Improving LIME Robustness with Smarter Locality Sampling
Autor: Saito, Sean, Chua, Eugene, Capel, Nicholas, Hu, Rocco
Explainability algorithms such as LIME have enabled machine learning systems to adopt transparency and fairness, which are important qualities in commercial use cases. However, recent work has shown that LIME's naive sampling strategy can be exploite
Externí odkaz: http://arxiv.org/abs/2006.12302
Zobrazit plný text záznamu
3
Report
Explaining Away Attacks Against Neural Networks
Autor: Saito, Sean, Wang, Jin
We investigate the problem of identifying adversarial attacks on image-based neural networks. We present intriguing experimental results showing significant discrepancies between the explanations generated for the predictions of a model on clean and
Externí odkaz: http://arxiv.org/abs/2003.05748
Zobrazit plný text záznamu
4
Report
Effects of Loss Functions And Target Representations on Adversarial Robustness
Autor: Saito, Sean, Roy, Sujoy
Understanding and evaluating the robustness of neural networks under adversarial settings is a subject of growing interest. Attacks proposed in the literature usually work with models trained to minimize cross-entropy loss and output softmax probabil
Externí odkaz: http://arxiv.org/abs/1812.00181
Zobrazit plný text záznamu
5
Report
HiNet: Hierarchical Classification with Neural Network
Autor: Wu, Zhenzhou, Saito, Sean
Traditionally, classifying large hierarchical labels with more than 10000 distinct traces can only be achieved with flatten labels. Although flatten labels is feasible, it misses the hierarchical information in the labels. Hierarchical models like HS
Externí odkaz: http://arxiv.org/abs/1705.11105
Zobrazit plný text záznamu
10
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.

Vyhledávací nástroje:

Upřesnit hledání

Omezení vyhledávání
Plný text Recenzováno Digitální knihovna AV ČR
Zdroje