Semantic Scholar Open Access 2021 68 sitasi

Firefly Neural Architecture Descent: a General Approach for Growing Neural Networks

Lemeng Wu Bo Liu P. Stone Qiang Liu

Abstrak

We propose firefly neural architecture descent, a general framework for progressively and dynamically growing neural networks to jointly optimize the networks' parameters and architectures. Our method works in a steepest descent fashion, which iteratively finds the best network within a functional neighborhood of the original network that includes a diverse set of candidate network structures. By using Taylor approximation, the optimal network structure in the neighborhood can be found with a greedy selection procedure. We show that firefly descent can flexibly grow networks both wider and deeper, and can be applied to learn accurate but resource-efficient neural architectures that avoid catastrophic forgetting in continual learning. Empirically, firefly descent achieves promising results on both neural architecture search and continual learning. In particular, on a challenging continual image classification task, it learns networks that are smaller in size but have higher average accuracy than those learned by the state-of-the-art methods.

Topik & Kata Kunci

Penulis (4)

L

Lemeng Wu

B

Bo Liu

P

P. Stone

Q

Qiang Liu

Format Sitasi

Wu, L., Liu, B., Stone, P., Liu, Q. (2021). Firefly Neural Architecture Descent: a General Approach for Growing Neural Networks. https://www.semanticscholar.org/paper/3e1b060ebacfc7a966ec735c940e2ee48f2a7a99

Akses Cepat

PDF tidak tersedia langsung

Cek di sumber asli →
Lihat di Sumber
Informasi Jurnal
Tahun Terbit
2021
Bahasa
en
Total Sitasi
68×
Sumber Database
Semantic Scholar
Akses
Open Access ✓