Three-dimensional feedforward neural networks and their realization by nano-devices

Authors:
Vlad P. Shmerko;Svetlana N. Yanushkevich
Affiliations:
Department of Electrical and Computer Engineering, University of Calgary;Department of Electrical and Computer Engineering, University of Calgary
Venue:
Artificial intelligence in logic design
Year:
2004

Citing 8
Cited 0

Topological Properties of Hypercubes

IEEE Transactions on Computers
Depth-Size Tradeoffs for Neural Computation

IEEE Transactions on Computers - Special issue on artificial neural networks
Spatial tessellations: concepts and applications of Voronoi diagrams

Spatial tessellations: concepts and applications of Voronoi diagrams
Incomplete hypercubes: embeddings of tree-related networks

Journal of Parallel and Distributed Computing
Handbook of Neural Computation

Handbook of Neural Computation
Ordered binary decision diagrams

Logic Synthesis and Verification
Silicon Single-Electron Devices and Their Applications

ISMVL '00 Proceedings of the 30th IEEE International Symposium on Multiple-Valued Logic
Representation of Logical Circuits by Linear Decision Diagrams with Extension to Nanostructures

Automation and Remote Control

Quantified Score

Hi-index	0.00

Visualization

Abstract

The three-dimensional (3D) model of a feedforward neural network (NN) based on so called N-hypercube topology is proposed. The N-hypercube is different from the classical hypercube used in communication theory, and in Boolean algebra. This new structure has been created based on a novel algorithm for embedding a binary decision tree and binary decision diagram into a N-hypercube. It is shown that N-hypercube topology is a reasonable solution to implement NN of threshold gates, in particular, on the single-electron devices. The 3D design methodology of feedforward NN is oriented to technology mapping to nanodevices. Results of extensive experimental study of feedforward networks consisting of over 3500 N-hypercubes are presented.