Generalizability Under Sensor Failure: Tokenization + Transformers Enable More Robust Latent Spaces
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866929282501050368 |
|---|---|
| author | Chau, Geeling An, Yujin Iqbal, Ahamed Raffey Chung, Soon-Jo Yue, Yisong Talukder, Sabera |
| author_facet | Chau, Geeling An, Yujin Iqbal, Ahamed Raffey Chung, Soon-Jo Yue, Yisong Talukder, Sabera |
| contents | A major goal in neuroscience is to discover neural data representations that generalize. This goal is challenged by variability along recording sessions (e.g. environment), subjects (e.g. varying neural structures), and sensors (e.g. sensor noise), among others. Recent work has begun to address generalization across sessions and subjects, but few study robustness to sensor failure which is highly prevalent in neuroscience experiments. In order to address these generalizability dimensions we first collect our own electroencephalography dataset with numerous sessions, subjects, and sensors, then study two time series models: EEGNet (Lawhern et al., 2018) and TOTEM (Talukder et al., 2024). EEGNet is a widely used convolutional neural network, while TOTEM is a discrete time series tokenizer and transformer model. We find that TOTEM outperforms or matches EEGNet across all generalizability cases. Finally through analysis of TOTEM's latent codebook we observe that tokenization enables generalization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_18546 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Generalizability Under Sensor Failure: Tokenization + Transformers Enable More Robust Latent Spaces Chau, Geeling An, Yujin Iqbal, Ahamed Raffey Chung, Soon-Jo Yue, Yisong Talukder, Sabera Machine Learning A major goal in neuroscience is to discover neural data representations that generalize. This goal is challenged by variability along recording sessions (e.g. environment), subjects (e.g. varying neural structures), and sensors (e.g. sensor noise), among others. Recent work has begun to address generalization across sessions and subjects, but few study robustness to sensor failure which is highly prevalent in neuroscience experiments. In order to address these generalizability dimensions we first collect our own electroencephalography dataset with numerous sessions, subjects, and sensors, then study two time series models: EEGNet (Lawhern et al., 2018) and TOTEM (Talukder et al., 2024). EEGNet is a widely used convolutional neural network, while TOTEM is a discrete time series tokenizer and transformer model. We find that TOTEM outperforms or matches EEGNet across all generalizability cases. Finally through analysis of TOTEM's latent codebook we observe that tokenization enables generalization. |
| title | Generalizability Under Sensor Failure: Tokenization + Transformers Enable More Robust Latent Spaces |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2402.18546 |