Recurrence Methods in the Analysis of Learning Processes

Abstract

The goal of most learning processes is to bring a machine into a set of “correct” states. In practice, however, it may be difficult to show that the process enters this target set. We present a condition that ensures that the process visits the target set infinitely often almost surely. This condition is easy to verify and is true for many well-known learning rules. To demonstrate the utility of this method, we apply it to four types of learning processes: the perceptron, learning rules governed by continuous energy functions, the Kohonen rule, and the committee machine.

Cite

Text

Mendelson and Nelken. "Recurrence Methods in the Analysis of Learning Processes." Neural Computation, 2001. doi:10.1162/08997660152469378

Markdown

[Mendelson and Nelken. "Recurrence Methods in the Analysis of Learning Processes." Neural Computation, 2001.](https://mlanthology.org/neco/2001/mendelson2001neco-recurrence/) doi:10.1162/08997660152469378

BibTeX

@article{mendelson2001neco-recurrence,
  title     = {{Recurrence Methods in the Analysis of Learning Processes}},
  author    = {Mendelson, Shahar and Nelken, Israel},
  journal   = {Neural Computation},
  year      = {2001},
  pages     = {1839-1861},
  doi       = {10.1162/08997660152469378},
  volume    = {13},
  url       = {https://mlanthology.org/neco/2001/mendelson2001neco-recurrence/}
}