Abstract | ||
---|---|---|
This paper deals with the problem of the relation between conventional computers and the neural networks approach. We concentrate our efforts on comparison of traditional computing elements flip-flops and their potential successors — neural network processing elements. We show that both flip-flops and artificial neurons exhibit very similar properties, although their goals, complexity and ways of computation are different. To compare flip-flops and neurons we use a universal theory for problem solving — the Calculus of Self-modifiable Algorithms. |
Year | DOI | Venue |
---|---|---|
1991 | 10.1007/3-540-54029-6_218 | ICCI |
Keywords | Field | DocType |
new generation,neural network processing elements,neural network | Flip,Sequential logic,Computer science,FLOPS,Artificial neuron,Time delay neural network,Artificial neural network,Computer engineering,Computation,Embedded system | Conference |
Volume | ISSN | ISBN |
497 | 0302-9743 | 3-540-54029-6 |
Citations | PageRank | References |
3 | 0.75 | 7 |
Authors | ||
1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Eugeniusz Eberbach | 1 | 38 | 8.70 |