Synaptic device for artificial neural network and artificial neural network

Abstract

The invention provides a synaptic device for an artificial neural network. The artificial neural network comprises multiple neurons which are connected through a synaptic device array, the synaptic device array comprises multiple synaptic devices, and two or three or more synaptic devices are connected in parallel. The invention further provides the artificial neural network including the synaptic devices.

Description

technical field [0001] The present invention relates to synaptic devices and artificial neural networks for artificial neural networks. Background technique [0002] An artificial neural network is a device that imitates a biological neural network. It applies machine learning algorithms and can be used to complete tasks such as recognition and classification of various types of information such as images and sounds. In general, an artificial neural network may include a plurality of neurons and an array of synaptic devices for interconnecting the neurons with each other. As an example, figure 1 Neuron 1 and the two synapse devices W1 and W2 connected thereto are schematically shown. respectively via the synaptic device W 1 and W 2 The other two neurons (not shown) connected to neuron 1 use X for the excitation signal of neuron 1 1 and x 2 express. Neuron 1 can be implemented, for example, with conventional CMOS circuits. Synaptic device W 1 and W 2 For example, it...

AI Technical Summary

Problems solved by technology

This will affect the reliability of the artificial neural network training, thereby reducing the accuracy of the artificial neural network at work

Images