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

Download:

Current browse context:

cs.LG

Change to browse by:

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: Butterfly-Net2: Simplified Butterfly-Net and Fourier Transform Initialization

Abstract: Structured CNN designed using the prior information of problems potentially improves efficiency over conventional CNNs in various tasks in solving PDEs and inverse problems in signal processing. This paper introduces BNet2, a simplified Butterfly-Net and inline with the conventional CNN. Moreover, a Fourier transform initialization is proposed for both BNet2 and CNN with guaranteed approximation power to represent the Fourier transform operator. Experimentally, BNet2 and the Fourier transform initialization strategy are tested on various tasks, including approximating Fourier transform operator, end-to-end solvers of linear and nonlinear PDEs in 1D, and denoising and deblurring of 1D signals. On all tasks, under the same initialization, BNet2 achieves similar accuracy as CNN but has fewer parameters. Fourier transform initialized BNet2 and CNN consistently improve the training and testing accuracy over the randomly initialized CNN.
Subjects: Machine Learning (cs.LG); Numerical Analysis (math.NA); Machine Learning (stat.ML)
Cite as: arXiv:1912.04154 [cs.LG]
  (or arXiv:1912.04154v1 [cs.LG] for this version)

Submission history

From: Yingzhou Li [view email]
[v1] Mon, 9 Dec 2019 16:25:32 GMT (617kb)

Link back to: arXiv, form interface, contact.