We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Simple initialization and parametrization of sinusoidal networks via their kernel bandwidth

Abstract: Neural networks with sinusoidal activations have been proposed as an alternative to networks with traditional activation functions. Despite their promise, particularly for learning implicit models, their training behavior is not yet fully understood, leading to a number of empirical design choices that are not well justified. In this work, we first propose a simplified version of such sinusoidal neural networks, which allows both for easier practical implementation and simpler theoretical analysis. We then analyze the behavior of these networks from the neural tangent kernel perspective and demonstrate that their kernel approximates a low-pass filter with an adjustable bandwidth. Finally, we utilize these insights to inform the sinusoidal network initialization, optimizing their performance for each of a series of tasks, including learning implicit models and solving differential equations.
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2211.14503 [cs.LG]
  (or arXiv:2211.14503v1 [cs.LG] for this version)

Submission history

From: Filipe de Avila Belbute-Peres [view email]
[v1] Sat, 26 Nov 2022 07:41:48 GMT (38943kb,D)

Link back to: arXiv, form interface, contact.