| Автор | K. Y. M. Wong |
| Автор | M. Bouten |
| Дата выпуска | 1991-10-07 |
| dc.description | We demonstrate that the principle of adaptation enables us to prescribe the appropriate performance function for optimal robustness against random dilution, in both the entire or restricted space of synaptic weights. It also enables us to improve the network robustness against clipping. However, we clarify that clipping and annealed dilution result in different optimal networks, although they have increasing resemblance at low dilution. |
| Формат | application.pdf |
| Издатель | Institute of Physics Publishing |
| Название | The Principle of Adaptation and Dilution Robustness in Neural Networks |
| Тип | lett |
| DOI | 10.1209/0295-5075/16/6/003 |
| Electronic ISSN | 1286-4854 |
| Print ISSN | 0295-5075 |
| Журнал | EPL (Europhysics Letters) |
| Том | 16 |
| Первая страница | 525 |
| Последняя страница | 530 |
| Выпуск | 6 |
