Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2504.05677 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908306575982592 |
|---|---|
| author | Sakai, Shunsuke Tsuge, Shunsuke Hasegawa, Tatsuhito |
| author_facet | Sakai, Shunsuke Tsuge, Shunsuke Hasegawa, Tatsuhito |
| contents | Neural network ensembles is a simple yet effective approach for enhancing generalization capabilities. The most common method involves independently training multiple neural networks initialized with different weights and then averaging their predictions during inference. However, this approach increases training time linearly with the number of ensemble members. To address this issue, we propose the novel ``\textbf{Noisy Deep Ensemble}'' method, significantly reducing the training time required for neural network ensembles. In this method, a \textit{parent model} is trained until convergence, and then the weights of the \textit{parent model} are perturbed in various ways to construct multiple \textit{child models}. This perturbation of the \textit{parent model} weights facilitates the exploration of different local minima while significantly reducing the training time for each ensemble member. We evaluated our method using diverse CNN architectures on CIFAR-10 and CIFAR-100 datasets, surpassing conventional efficient ensemble methods and achieving test accuracy comparable to standard ensembles. Code is available at \href{https://github.com/TSTB-dev/NoisyDeepEnsemble}{https://github.com/TSTB-dev/NoisyDeepEnsemble} |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_05677 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Noisy Deep Ensemble: Accelerating Deep Ensemble Learning via Noise Injection Sakai, Shunsuke Tsuge, Shunsuke Hasegawa, Tatsuhito Computer Vision and Pattern Recognition Neural network ensembles is a simple yet effective approach for enhancing generalization capabilities. The most common method involves independently training multiple neural networks initialized with different weights and then averaging their predictions during inference. However, this approach increases training time linearly with the number of ensemble members. To address this issue, we propose the novel ``\textbf{Noisy Deep Ensemble}'' method, significantly reducing the training time required for neural network ensembles. In this method, a \textit{parent model} is trained until convergence, and then the weights of the \textit{parent model} are perturbed in various ways to construct multiple \textit{child models}. This perturbation of the \textit{parent model} weights facilitates the exploration of different local minima while significantly reducing the training time for each ensemble member. We evaluated our method using diverse CNN architectures on CIFAR-10 and CIFAR-100 datasets, surpassing conventional efficient ensemble methods and achieving test accuracy comparable to standard ensembles. Code is available at \href{https://github.com/TSTB-dev/NoisyDeepEnsemble}{https://github.com/TSTB-dev/NoisyDeepEnsemble} |
| title | Noisy Deep Ensemble: Accelerating Deep Ensemble Learning via Noise Injection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.05677 |