On-line AdaTron learning of unlearnable rules

Jun ichi Inoue, Hidetoshi Nishimori

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)


We study the on-line AdaTron learning of linearly nonseparable rules by a simple perceptron. Training examples are provided by a perceptron with a nonmonotonic transfer function that reduces to the usual monotonic relation in a certain limit. We find that, although the on-line AdaTron learning is a powerful algorithm for the learnable rule, it does not give the best possible generalization error for unlearnable problems. Optimization of the learning rate is shown to greatly improve the performance of the AdaTron algorithm, leading to the best possible generalization error for a wide range of the parameter that controls the shape of the transfer function.

Original languageEnglish
Pages (from-to)4544-4551
Number of pages8
JournalPhysical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics
Issue number4
Publication statusPublished - 1997
Externally publishedYes

ASJC Scopus subject areas

  • Statistical and Nonlinear Physics
  • Statistics and Probability
  • Condensed Matter Physics


Dive into the research topics of 'On-line AdaTron learning of unlearnable rules'. Together they form a unique fingerprint.

Cite this