Function-space Parameterization of Neural Networks for Sequential Learning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914717397680128 |
|---|---|
| author | Scannell, Aidan Mereu, Riccardo Chang, Paul Tamir, Ella Pajarinen, Joni Solin, Arno |
| author_facet | Scannell, Aidan Mereu, Riccardo Chang, Paul Tamir, Ella Pajarinen, Joni Solin, Arno |
| contents | Sequential learning paradigms pose challenges for gradient-based deep learning due to difficulties incorporating new data and retaining prior knowledge. While Gaussian processes elegantly tackle these problems, they struggle with scalability and handling rich inputs, such as images. To address these issues, we introduce a technique that converts neural networks from weight space to function space, through a dual parameterization. Our parameterization offers: (i) a way to scale function-space methods to large data sets via sparsification, (ii) retention of prior knowledge when access to past data is limited, and (iii) a mechanism to incorporate new data without retraining. Our experiments demonstrate that we can retain knowledge in continual learning and incorporate new data efficiently. We further show its strengths in uncertainty quantification and guiding exploration in model-based RL. Further information and code is available on the project website. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_10929 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Function-space Parameterization of Neural Networks for Sequential Learning Scannell, Aidan Mereu, Riccardo Chang, Paul Tamir, Ella Pajarinen, Joni Solin, Arno Machine Learning Sequential learning paradigms pose challenges for gradient-based deep learning due to difficulties incorporating new data and retaining prior knowledge. While Gaussian processes elegantly tackle these problems, they struggle with scalability and handling rich inputs, such as images. To address these issues, we introduce a technique that converts neural networks from weight space to function space, through a dual parameterization. Our parameterization offers: (i) a way to scale function-space methods to large data sets via sparsification, (ii) retention of prior knowledge when access to past data is limited, and (iii) a mechanism to incorporate new data without retraining. Our experiments demonstrate that we can retain knowledge in continual learning and incorporate new data efficiently. We further show its strengths in uncertainty quantification and guiding exploration in model-based RL. Further information and code is available on the project website. |
| title | Function-space Parameterization of Neural Networks for Sequential Learning |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2403.10929 |