Stochastic deep learning in memristive network s

Anakha V. Babu, Bipin Rajendran

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations

Abstract

We study the performance of stochastically trained deep neural networks (DNNs) whose synaptic weights are implemented using emerging memristive devices that exhibit limited dynamic range, resolution, and variability in their programming characteristics. We show that a key device parameter to optimize the learning efficiency of DNNs is the variability in its programming characteristics. DNNs with such memristive synapses, even with dynamic range as low as 15 and only 32 discrete levels, when trained based on stochastic updates suffer less than 3% loss in accuracy compared to floating point software baseline. We also study the performance of stochastic memristive DNNs when used as inference engines with noise corrupted data and find that if the device variability can be minimized, the relative degradation in performance for the Stochastic DNN is better than that of the software baseline. Hence, our study presents a new optimization corner for memristive devices for building large noise-immune deep learning systems.

Original languageEnglish (US)
Title of host publicationICECS 2017 - 24th IEEE International Conference on Electronics, Circuits and Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages214-217
Number of pages4
ISBN (Electronic)9781538619117
DOIs
StatePublished - Feb 14 2018
Event24th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2017 - Batumi, Georgia
Duration: Dec 5 2017Dec 8 2017

Publication series

NameICECS 2017 - 24th IEEE International Conference on Electronics, Circuits and Systems
Volume2018-January

Other

Other24th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2017
Country/TerritoryGeorgia
CityBatumi
Period12/5/1712/8/17

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Energy Engineering and Power Technology

Fingerprint

Dive into the research topics of 'Stochastic deep learning in memristive network s'. Together they form a unique fingerprint.

Cite this