paper-with-me

Papers

If dropout limits trainable depth, does critical initialisation still matter? A large-scale statistical analysis on ReLU networks

2019-10-13 · Arnu Pretorius, Elan van Biljon, Benjamin van Niekerk, Ryan Eloff, Matthew Reynard, Steve James, Benjamin Rosman, Herman Kamper, Steve Kroon

Recent work in signal propagation theory has shown that dropout limits the depth to which information can propagate through a neural network. In this paper, we investigate the effect of initialisation on training speed and generalisation for ReLU networks within this depth limit. We ask the following research question: given that critical initialisation is crucial for training at large depth, if dropout limits the depth at which networks are trainable, does initialising critically still matter? We conduct a large-scale controlled experiment, and perform a statistical analysis of over $12000$ trained networks. We find that (1) trainable networks show no statistically significant difference in performance over a wide range of non-critical initialisations; (2) for initialisations that show a statistically significant difference, the net effect on performance is small; (3) only extreme initialisations (very small or very large) perform worse than criticality. These findings also apply to standard ReLU networks of moderate depth as a special case of zero dropout. Our results therefore suggest that, in the shallow-to-moderate depth setting, critical initialisation provides zero performance gains when compared to off-critical initialisations and that searching for off-critical initialisations that might improve training speed or generalisation, is likely to be a fruitless endeavour.

📄 PDF Abstract BibTeX arXiv:1910.05725

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…
ReLU How Do I Communicate to Expedia? How Do I Communicate to Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Live Support & Special Travel…
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…

Similar Papers 제목 키워드 기반

Deep Information Propagation

2016-11-04 · Samuel S. Schoenholz, Justin Gilmer, Surya Ganguli, Jascha Sohl-Dickstein

We study the behavior of untrained neural networks whose weights and biases are randomly distributed using mean field theory. We show the existence of depth scales that naturally limit the maximum depth of signal propaga…

Mean field theory for deep dropout networks: digging up gradient backpropagation deeply

2019-12-19 · Wei Huang, Richard Yi Da Xu, Weitao Du, Yutian Zeng 외

In recent years, the mean field theory has been applied to the study of neural networks and has achieved a great deal of success. The theory has been applied to various neural network structures, including CNNs, RNNs, Re…

Self-Knowledge Distillation via Dropout

2022-08-11 · Hyoje Lee, Yeachan Park, Hyun Seo, Myungjoo Kang

To boost the performance, deep neural networks require deeper or wider network structures that involve massive computational and memory costs. To alleviate this issue, the self-knowledge distillation method regularizes t…

Adversarial Robustnessimage-classificationImage ClassificationKnowledge Distillation+4

Critical initialisation for deep signal propagation in noisy rectifier neural networks

2018-11-01 · NeurIPS 2018 12 · Arnu Pretorius, Elan van Biljon, Steve Kroon, Herman Kamper

Stochastic regularisation is an important weapon in the arsenal of a deep learning practitioner. However, despite recent theoretical advances, our understanding of how noise influences signal propagation in deep neural n…

Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos

2026-05-20 · Lucas Fernandez Sarmiento arxiv

We develop a mean-field theory of dropout as a perturbation of critical signal propagation at the edge of chaos, and show that it predicts a simple, no-cost change to standard practice: \emph{front-loaded} dropout schedu…