Constructive Learning Using Internal Representation Conflicts

Abstract

We present an algorithm for the training of feedforward and recur(cid:173) rent neural networks. It detects internal representation conflicts and uses these conflicts in a constructive manner to add new neu(cid:173) rons to the network . The advantages are twofold: (1) starting with a small network neurons are only allocated when required; (2) by detecting and resolving internal conflicts at an early stage learning time is reduced. Empirical results on two real-world problems sub(cid:173) stantiate the faster learning speed; when applied to the training of a recurrent network on a well researched sequence recognition task (the Reber grammar), training times are significantly less than previously reported .

Cite

Text

Leerink and Jabri. "Constructive Learning Using Internal Representation Conflicts." Neural Information Processing Systems, 1993.

Markdown

[Leerink and Jabri. "Constructive Learning Using Internal Representation Conflicts." Neural Information Processing Systems, 1993.](https://mlanthology.org/neurips/1993/leerink1993neurips-constructive/)

BibTeX

@inproceedings{leerink1993neurips-constructive,
  title     = {{Constructive Learning Using Internal Representation Conflicts}},
  author    = {Leerink, Laurens R. and Jabri, Marwan A.},
  booktitle = {Neural Information Processing Systems},
  year      = {1993},
  pages     = {279-284},
  url       = {https://mlanthology.org/neurips/1993/leerink1993neurips-constructive/}
}