This paper describes a memristor-based neuromorphic system that can be used for ex-situ training of various multi-layer neural network algorithms. This system is based on an analog neuron circuit that is capable of performing an accurate dot product calculation. The presented ex-situ programming technique can be used to map many key neural algorithms directly onto the grid of resistances in a memristor crossbar. Using this weight-to-crossbar mapping approach along with the dot product calculation circuit, complex neural algorithms can be easily implemented using this system. To show the effectiveness and versatility of this circuit, a Multilayer Perceptron (MLP) is trained to perform Sobel edge detection. Following these simulations, an analysis was presented that shows how both memristor accuracy and neuron circuit gain relates to output error. Additionally, this paper discusses how circuit noise and neural network layout contribute to testing accuracy.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Memristor crossbar based implementation of a multilayer perceptron


    Contributors:


    Publication date :

    2017-06-01


    Size :

    521715 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Hybrid crossbar architecture for a memristor based memory

    Yakopcic, Chris / Taha, Tarek M. / Hasan, Raqibul | IEEE | 2014



    Cognitive domain ontologies in a memristor crossbar architecture

    Yakopcic, Chris / Rahman, Nayim / Atahary, Tanvir et al. | IEEE | 2017


    Evaluation of Leakage Currents in Memristor Crossbar Arrays

    Kasongo, Muana / Ytterdal, Trond / Lee, John et al. | IEEE | 2023


    Modular, Multilayer Perceptron

    Cheng, Li-Jen / Liu, Tsuen-Hsi | NTRS | 1991