Bala, Anu ORCID: https://orcid.org/0009-0000-6242-5248, Yang, Xiaohan, Adeyemo, Adedotun, Khandelwal, Saurabh and jabir, Abusaleh (2024) Memristor crossbar-based learning method for ex situ training in neural networks. In: Nanoscale Memristor Device and Circuits Design. Elsevier

Full text not available from this repository.

Abstract

Memristors are being considered as a game changer for the realization of neuromorphic hardware systems due to their similarity to biological synapses. Recent studies show that memristor crossbar arrays can provide high-density and high-performance neural-network hardware implementation at low power due to their physical layout, nanoscale size, and low power consumption features. This chapter describes a training method that can be used for the implementation of memristive multilayer neural networks using an ex situ method. We mimic the behavior of memristor crossbars in a software training process to achieve more accurate and close computations than with hardware. A voltage divider has been used to calculate the dot product in this method. To demonstrate the accuracy and effectiveness of this method, different patterns and nonseparable functions using memristor crossbar structures are simulated. The results demonstrate that more accurate computations can be produced using this learning method for ex situ training. It also reduces the learning time of functions.

Item Type:	Book Section
Status:	Published
DOI:	10.1016/b978-0-323-90793-4.00009-x
Subjects:	T Technology > T Technology (General)
School/Department:	London Campus
URI:	https://ray.yorksj.ac.uk/id/eprint/13534

University Staff: Request a correction | RaY Editors: Update this record