The problem of construction of idealizing sequence of spikes as the weighted sum of input sequences of spikes, which is important for learning procedure of spike neural network on the basis of methods of linear algebra with the use of idealizing impulses with a zero duration and single amplitude and conception of space of sequences of spikes is considered. For determination of weight coefficients a least squares method is used. Resultant spike sequence is formed with the use of fuzzy numbers. Two examples of approximation of the sequence of spikes are demonstrated.
1. Maass W. Networks of spiking neurons: The third generation of neural network models / W. Maass // Neural Networks — 1997. — Vol. 10, No. 9. — P. 1659–1671.
2. Бодянский Е.В. Гетерогенная спайк-нейронная сеть с латеральными связями в задаче кластеризации / Е.В. Бодянский, А.И. Долотов // Системи обробки інформації. — 2007. — Вип. 8 (66). — С. 10–15.
3. Kasinski A. Comparison of Supervised Learning Methods for Spike Time Coding in Spiking Neural Networks / A. Kasinski, F. Ponulak // Int. J. Appl. Math. Comput. Sci. — 2006. — Vol. 16, No. 1. — P. 101–113.
4. Carnell A. Linear Algebra for Time Series of Spikes [Електронний ресурс] / A. Carnell, D. Richardson // University of Bath, UK. —
2002. — 7 p. — Режим доступу: esann.pdf.
5. Корн Г. Справочник по математике для научных работников и инженеров / Г. Корн, Т. Корн. — М.: Наука, 1970. — 720 с.