Natural Gradient Descent for Training Stochastic Complex-Valued Neural Networks

Abstract

In this paper, the natural gradient descent method for the multilayer stochastic complex-valued neural networks is considered, and the natural gradient is given for a single stochastic complex-valued neuron as an example. Since the space of the learnable parameters of stochastic complex-valued neural networks is not the Euclidean space but a curved manifold, the complex-valued natural gradient method is expected to exhibit excellent learning performance.

Authors and Affiliations

Tohru Nitta

Keywords

Related Articles

Enhanced Mechanism to Detect and Mitigate Economic Denial of Sustainability (EDoS) Attack in Cloud Computing Environments

Cloud computing (CC) is the next revolution in the Information and Communication Technology arena. CC is often provided as a service comparable to utility services such as electricity, water, and telecommunications. Clou...

  Enhancement of Passive MAC Spoofing Detection Techniques

 Failure of addressing all IEEE 802.11i Robust Security Networks (RSNs) vulnerabilities enforces many researchers to revise robust and reliable Wireless Intrusion Detection Techniques (WIDTs). In this paper we propo...

TokenSign: Using Revocable Fingerprint Biotokens and Secret Sharing Scheme as Electronic Signature

Electronic signature is a quick and convenient tool, used for legal documents and payments since business practices revolutionized from traditional paper-based to computer-based systems. The growing use of electronic sig...

A Case Study for the IONEX CODE-Database Processing Tool Software: Ionospheric Anomalies before the Mw 8.2 Earthquake in Mexico on September 7, 2017

A software tool was developed in the Imaging Processing Research laboratory (INTI-Lab) that automatically downloads several IONEX files around a specific user input date and also performs statistical calculations to look...

A Topic Modeling Based Solution for Confirming Software Documentation Quality

this paper presents an approach for evaluating and confirming the quality of the external software documentation using topic modeling. Typically, the quality of the external documentation has to mirror precisely the orga...

Download PDF file
  • EP ID EP121478
  • DOI 10.14569/IJACSA.2014.050729
  • Views 73
  • Downloads 0

How To Cite

Tohru Nitta (2014). Natural Gradient Descent for Training Stochastic Complex-Valued Neural Networks. International Journal of Advanced Computer Science & Applications, 5(7), 193-198. https://europub.co.uk/articles/-A-121478