Premium
A Ballistic Transport Model for an Artificial Neuron
Author(s) -
Nemnes George Alexandru,
Dragoman Daniela
Publication year - 2020
Publication title -
physica status solidi (a)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.532
H-Index - 104
eISSN - 1862-6319
pISSN - 1862-6300
DOI - 10.1002/pssa.201900936
Subject(s) - artificial neural network , function (biology) , scattering , transmission (telecommunications) , ballistic conduction , voltage , process (computing) , set (abstract data type) , amplitude , activation function , computer science , matrix (chemical analysis) , physics , topology (electrical circuits) , control theory (sociology) , mathematics , artificial intelligence , quantum mechanics , materials science , telecommunications , combinatorics , evolutionary biology , programming language , composite material , operating system , electron , control (management) , biology
A model for an artificial neuron (AN) is introduced which is based on ballistic transport in a multiterminal device. Unlike standard configurations, the proposed design embeds the synaptic weights into the active region, thus significantly reducing the complexity of the input terminals. This is achieved by defining the basic elements of the ballistic AN as follows: the input values are set by the incoming wavefunctions amplitudes, whereas the weights correspond to the scattering matrix elements. Furthermore, the output value of the activation function of the AN is given by the transmission function. By tuning the gate voltage, the scattering potential and, consequently, the weights are changed so that the value of the transmission function gets closer to the target output, which is essential in the training process of artificial neural networks. Thus, the modus operandi of a ballistic AN is provided.