Correlation Analysis of Adversarial Attack in Time Series Classification
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910571728732160 |
|---|---|
| author | Li, Zhengyang Liang, Wenhao Dong, Chang Chen, Weitong Huang, Dong |
| author_facet | Li, Zhengyang Liang, Wenhao Dong, Chang Chen, Weitong Huang, Dong |
| contents | This study investigates the vulnerability of time series classification models to adversarial attacks, with a focus on how these models process local versus global information under such conditions. By leveraging the Normalized Auto Correlation Function (NACF), an exploration into the inclination of neural networks is conducted. It is demonstrated that regularization techniques, particularly those employing Fast Fourier Transform (FFT) methods and targeting frequency components of perturbations, markedly enhance the effectiveness of attacks. Meanwhile, the defense strategies, like noise introduction and Gaussian filtering, are shown to significantly lower the Attack Success Rate (ASR), with approaches based on noise introducing notably effective in countering high-frequency distortions. Furthermore, models designed to prioritize global information are revealed to possess greater resistance to adversarial manipulations. These results underline the importance of designing attack and defense mechanisms, informed by frequency domain analysis, as a means to considerably reinforce the resilience of neural network models against adversarial threats. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_11264 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Correlation Analysis of Adversarial Attack in Time Series Classification Li, Zhengyang Liang, Wenhao Dong, Chang Chen, Weitong Huang, Dong Machine Learning Cryptography and Security I.2.0 This study investigates the vulnerability of time series classification models to adversarial attacks, with a focus on how these models process local versus global information under such conditions. By leveraging the Normalized Auto Correlation Function (NACF), an exploration into the inclination of neural networks is conducted. It is demonstrated that regularization techniques, particularly those employing Fast Fourier Transform (FFT) methods and targeting frequency components of perturbations, markedly enhance the effectiveness of attacks. Meanwhile, the defense strategies, like noise introduction and Gaussian filtering, are shown to significantly lower the Attack Success Rate (ASR), with approaches based on noise introducing notably effective in countering high-frequency distortions. Furthermore, models designed to prioritize global information are revealed to possess greater resistance to adversarial manipulations. These results underline the importance of designing attack and defense mechanisms, informed by frequency domain analysis, as a means to considerably reinforce the resilience of neural network models against adversarial threats. |
| title | Correlation Analysis of Adversarial Attack in Time Series Classification |
| topic | Machine Learning Cryptography and Security I.2.0 |
| url | https://arxiv.org/abs/2408.11264 |