Adversarial training using meta-learning for BERT
Deep learning is currently the most successful method of semantic analysis in natural language processing. However, in recent years, many variants of carefully crafted inputs designed to cause misclassification, known as adversarial attacks, have been engineered with tremendous success. One well-...
Main Author: | Low, Timothy Jing Haen |
---|---|
Other Authors: | Joty Shafiq Rayhan |
Format: | Final Year Project (FYP) |
Language: | English |
Published: |
Nanyang Technological University
2022
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/156635 |
Similar Items
-
Adversarial robustness of deep reinforcement learning
by: Qu, Xinghua
Published: (2022) -
Meta-learning for deep reinforcement learning
by: Poon, Jun Yaw
Published: (2021) -
Adversarial patch detection
by: Yeong, Joash Ler Yuen
Published: (2022) -
Defence on unrestricted adversarial examples
by: Chan, Jarod Yan Cheng
Published: (2021) -
Investigating the causes of the vulnerability of CNNs to adversarial perturbations: learning objective, model components, and learned representations
by: Coppola, Davide
Published: (2023)