The development of neural network models requires the study o] dedicated hardware architectures. In this paper, we propose an implementation of Radial Basis Function networks, derive an architecture based on an already existing 2D-systolie machine (MANTRA). A systolic algorithm is described to implement the required functions and the suitable sequence of operations. Theoretical effieiencies are estimated on the key tasks and some guidelines ore given for a best usage of the Mantra machine in the studied Framework.
CITATION STYLE
Blayo, F., Guèrin-Duguè, A., & Maria, N. (1995). Implementing radial basis functions neural networks on the systolic MANTRA machine. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 930, pp. 781–788). Springer Verlag. https://doi.org/10.1007/3-540-59497-3_250
Mendeley helps you to discover research relevant for your work.