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 > Information Retrieval

Title: Parametric Rectified Power Sigmoid Units: Learning Nonlinear Neural Transfer Analytical Forms

Abstract: The paper proposes representation functionals in a dual paradigm where learning jointly concerns both linear convolutional weights and parametric forms of nonlinear activation functions. The nonlinear forms proposed for performing the functional representation are associated with a new class of parametric neural transfer functions called rectified power sigmoid units. This class is constructed to integrate both advantages of sigmoid and rectified linear unit functions, in addition with rejecting the drawbacks of these functions. Moreover, the analytic form of this new neural class involves scale, shift and shape parameters so as to obtain a wide range of activation shapes, including the standard rectified linear unit as a limit case. Parameters of this neural transfer class are considered as learnable for the sake of discovering the complex shapes that can contribute in solving machine learning issues. Performance achieved by the joint learning of convolutional and rectified power sigmoid learnable parameters are shown outstanding in both shallow and deep learning frameworks. This class opens new prospects with respect to machine learning in the sense that learnable parameters are not only attached to linear transformations, but also to suitable nonlinear operators.
Subjects: Information Retrieval (cs.IR); Machine Learning (cs.LG); Signal Processing (eess.SP); Functional Analysis (math.FA)
Cite as: arXiv:2101.09948 [cs.IR]
  (or arXiv:2101.09948v2 [cs.IR] for this version)

Submission history

From: Abdourrahmane Mahamane Atto [view email]
[v1] Mon, 25 Jan 2021 08:25:22 GMT (2248kb,D)
[v2] Fri, 5 Feb 2021 11:42:08 GMT (3077kb,D)

Link back to: arXiv, form interface, contact.