ML Anthology

Recovering a Feed-Forward Net from Its Output

Charles Fefferman, Scott Markel
NeurIPS 1993 pp. 335-342
/neurips/1993/fefferman1993neurips-recovering/

Abstract

We study feed-forward nets with arbitrarily many layers, using the stan(cid:173) dard sigmoid, tanh x. Aside from technicalities, our theorems are: 1. Complete knowledge of the output of a neural net for arbitrary inputs uniquely specifies the architecture, weights and thresholds; and 2. There are only finitely many critical points on the error surface for a generic training problem.

PDF NeurIPS Semantic Scholar

Cite

Text

Fefferman and Markel. "Recovering a Feed-Forward Net from Its Output." Neural Information Processing Systems, 1993.

Markdown

[Fefferman and Markel. "Recovering a Feed-Forward Net from Its Output." Neural Information Processing Systems, 1993.](https://mlanthology.org/neurips/1993/fefferman1993neurips-recovering/)

BibTeX

@inproceedings{fefferman1993neurips-recovering,
  title     = {{Recovering a Feed-Forward Net from Its Output}},
  author    = {Fefferman, Charles and Markel, Scott},
  booktitle = {Neural Information Processing Systems},
  year      = {1993},
  pages     = {335-342},
  url       = {https://mlanthology.org/neurips/1993/fefferman1993neurips-recovering/}
}