A LEARNING ALGORITHM USING SPECTRUM THEORY

  • Sonohara Hikaru
    Systems & Software Engineering Laboratory, Research & Development Center, Toshiba Co.
  • Tsukimoto Hiroshi
    Systems & Software Engineering Laboratory, Research & Development Center, Toshiba Co.

Bibliographic Information

Other Title
  • スペクトル理論を用いた学習アルゴリズムについて

Search this article

Description

As learning and prediction methods from numerical data, for example, linear regression analysis has been used. We have been studying non-linear regression analysis algorithms. In 1993, Linial presented a learning algorithm based on spectrum theory (Fourier analysis for Boolean functions). In this paper, we extend the spectrum theory and the learning algorithm to continuous domains in order to deal with numerical data. We experimentally confirm that the extended algorithm works well compared with Linial's algorithm.

Journal

Details 詳細情報について

  • CRID
    1571417127079861120
  • NII Article ID
    110002674888
  • NII Book ID
    AN10464118
  • ISSN
    09196072
  • Text Lang
    ja
  • Data Source
    • CiNii Articles

Report a problem

Back to top