Natural Gradient Works Efficiently in Learning

Article Properties
  • Language
    English
  • Publication Date
    1998/02/01
  • Indian UGC (Journal)
  • Refrences
    22
  • Citations
    922
  • Shun-ichi Amari RIKEN Frontier Research Program, Saitama 351-01, Japan
Abstract
Cite
Amari, Shun-ichi. “Natural Gradient Works Efficiently in Learning”. Neural Computation, vol. 10, no. 2, 1998, pp. 251-76, https://doi.org/10.1162/089976698300017746.
Amari, S.- ichi. (1998). Natural Gradient Works Efficiently in Learning. Neural Computation, 10(2), 251-276. https://doi.org/10.1162/089976698300017746
Amari S ichi. Natural Gradient Works Efficiently in Learning. Neural Computation. 1998;10(2):251-76.
Journal Categories
Medicine
Internal medicine
Neurosciences
Biological psychiatry
Neuropsychiatry
Science
Mathematics
Instruments and machines
Electronic computers
Computer science
Technology
Electrical engineering
Electronics
Nuclear engineering
Electronics
Technology
Mechanical engineering and machinery
Description

Is the natural gradient the key to efficient learning in complex systems? This research explores the advantages of using the natural gradient, rather than the ordinary gradient, for learning in parameter spaces with underlying structures. The study emphasizes that the ordinary gradient may not accurately represent the steepest direction of a function in such spaces, while the natural gradient does. The dynamics are analyzed, and an adaptive method for updating the learning rate is proposed. Information geometry is employed to calculate natural gradients in various contexts, including the parameter space of perceptrons, matrices (for blind source separation), and linear dynamical systems (for blind source deconvolution). These calculations provide a theoretical foundation for the method's effectiveness. The learning rate can be updated using an adaptive method. Through analysis, the natural gradient online learning is shown to be Fisher efficient, implying asymptotically optimal performance comparable to batch estimation. This suggests that the plateau phenomenon often observed in backpropagation learning algorithms may be mitigated by using the natural gradient. The research offers a valuable approach for improving the efficiency of learning in complex systems.

Published in Neural Computation, this paper addresses a core topic in the field of neural networks and machine learning. The journal focuses on computational and theoretical aspects of brain function and intelligent systems, making this exploration of the natural gradient highly relevant. The research contributes to the ongoing development of more efficient learning algorithms, a central theme of the journal.

Refrences
Citations
Citations Analysis
The first research to cite this article was titled A First Application of Independent Component Analysis to Extracting Structure from Stock Returns and was published in 1997. The most recent citation comes from a 2024 study titled A First Application of Independent Component Analysis to Extracting Structure from Stock Returns . This article reached its peak citation in 2022 , with 78 citations.It has been cited in 349 different journals, 14% of which are open access. Among related journals, the Neurocomputing cited this research the most, with 52 citations. The chart below illustrates the annual citation trends for this article.
Citations used this article by year