350 rub
Journal Neurocomputers №5 for 2010 г.
Article in number:
Fault-tolerant functionally redundant artificial neural networks with individual adaptation to neurons faults
Authors:
I. V. Potapov
Abstract:
The approaches of fault-tolerant functionally redundant and structurally uniform neural networks developing are considered. The adaptation to faults is realized by setting the values of multifunctional neurons. Criterion of system reliability and com-plexity ratio is produced. Probabilistic model of system working for estimation the reliability measures is derived. Some exam-ples are examined.
Pages: 3-10
References
- Острейковский В. А. Теория надежности. М.: Высшая школа. 2003.
- Мкртчян С. О. Нейроны и нейронные сети. М.: Энергия. 1971.
- Потапов И. В. Надежность нейрокомпьютерных систем. Модели и задачи. Омск: ОмГТУ. 2007.
- Половко А. М., Гуров С. В. Основы теории надежности. 2-е изд., перераб. и доп. СПб.: БХВ-Петербург. 2006.