DocumentCode
3224427
Title
SET based Boltzmann machine and Hopfield neural networks
Author
Liu, Chia-Chin ; Chen, Chunhong
Author_Institution
Dept. of Electr. & Comput. Eng., Univ. of Windsor, Windsor, ON, Canada
fYear
2011
fDate
15-18 Aug. 2011
Firstpage
413
Lastpage
416
Abstract
This paper presents a method of implementing Boltzmann machine and Hopfield neural networks using single electron devices. Comparison between these two networks is shown by computer simulations in terms of their ability to converge to a global minimum state. It is demonstrated that the probabilistic nature of single electron tunneling phenomena enables the stochastic neuron operation with Boltzmann machine.
Keywords
Boltzmann machines; Hopfield neural nets; single electron devices; stochastic processes; Boltzmann machine; Hopfield neural networks; SET; computer simulations; single electron devices; single electron tunneling; stochastic neuron operation; Biological neural networks; Capacitance; Energy states; Inverters; Neurons; Simulation; Stochastic processes; Boltzmann machine; Hopfield neural network; global minimum energy state; single-electron tunnelling;
fLanguage
English
Publisher
ieee
Conference_Titel
Nanotechnology (IEEE-NANO), 2011 11th IEEE Conference on
Conference_Location
Portland, OR
ISSN
1944-9399
Print_ISBN
978-1-4577-1514-3
Electronic_ISBN
1944-9399
Type
conf
DOI
10.1109/NANO.2011.6144315
Filename
6144315
Link To Document