Committee machines—a universal method to deal with non-idealities in memristor-based neural networks
Designing reliable and energy-efficient memristor-based artificial neural networks remains a challenge. Here, the authors demonstrate a technology-agnostic approach, committee machines, which increases the inference accuracy of memristive neural networks that suffer from device variability, faulty d...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2020-08-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-020-18098-0 |
| _version_ | 1818421194424057856 |
|---|---|
| author | D. Joksas P. Freitas Z. Chai W. H. Ng M. Buckwell C. Li W. D. Zhang Q. Xia A. J. Kenyon A. Mehonic |
| author_facet | D. Joksas P. Freitas Z. Chai W. H. Ng M. Buckwell C. Li W. D. Zhang Q. Xia A. J. Kenyon A. Mehonic |
| author_sort | D. Joksas |
| collection | DOAJ |
| description | Designing reliable and energy-efficient memristor-based artificial neural networks remains a challenge. Here, the authors demonstrate a technology-agnostic approach, committee machines, which increases the inference accuracy of memristive neural networks that suffer from device variability, faulty devices, random telegraph noise and line resistance. |
| first_indexed | 2024-12-14T13:06:29Z |
| format | Article |
| id | doaj.art-1568419195894594b620bd5c234eada1 |
| institution | Directory Open Access Journal |
| issn | 2041-1723 |
| language | English |
| last_indexed | 2024-12-14T13:06:29Z |
| publishDate | 2020-08-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj.art-1568419195894594b620bd5c234eada12022-12-21T23:00:18ZengNature PortfolioNature Communications2041-17232020-08-0111111010.1038/s41467-020-18098-0Committee machines—a universal method to deal with non-idealities in memristor-based neural networksD. Joksas0P. Freitas1Z. Chai2W. H. Ng3M. Buckwell4C. Li5W. D. Zhang6Q. Xia7A. J. Kenyon8A. Mehonic9Department of Electronic and Electrical Engineering, University College London, Roberts Building, Torrington PlaceDepartment of Electronics and Electrical Engineering, Liverpool John Moores University, Liverpool, James Parsons Building, Byrom StreetDepartment of Electronics and Electrical Engineering, Liverpool John Moores University, Liverpool, James Parsons Building, Byrom StreetDepartment of Electronic and Electrical Engineering, University College London, Roberts Building, Torrington PlaceDepartment of Electronic and Electrical Engineering, University College London, Roberts Building, Torrington PlaceDepartment of Electrical and Computer Engineering, University of Massachusetts Amherst, 100 Natural Resources RoadDepartment of Electronics and Electrical Engineering, Liverpool John Moores University, Liverpool, James Parsons Building, Byrom StreetDepartment of Electrical and Computer Engineering, University of Massachusetts Amherst, 100 Natural Resources RoadDepartment of Electronic and Electrical Engineering, University College London, Roberts Building, Torrington PlaceDepartment of Electronic and Electrical Engineering, University College London, Roberts Building, Torrington PlaceDesigning reliable and energy-efficient memristor-based artificial neural networks remains a challenge. Here, the authors demonstrate a technology-agnostic approach, committee machines, which increases the inference accuracy of memristive neural networks that suffer from device variability, faulty devices, random telegraph noise and line resistance.https://doi.org/10.1038/s41467-020-18098-0 |
| spellingShingle | D. Joksas P. Freitas Z. Chai W. H. Ng M. Buckwell C. Li W. D. Zhang Q. Xia A. J. Kenyon A. Mehonic Committee machines—a universal method to deal with non-idealities in memristor-based neural networks Nature Communications |
| title | Committee machines—a universal method to deal with non-idealities in memristor-based neural networks |
| title_full | Committee machines—a universal method to deal with non-idealities in memristor-based neural networks |
| title_fullStr | Committee machines—a universal method to deal with non-idealities in memristor-based neural networks |
| title_full_unstemmed | Committee machines—a universal method to deal with non-idealities in memristor-based neural networks |
| title_short | Committee machines—a universal method to deal with non-idealities in memristor-based neural networks |
| title_sort | committee machines a universal method to deal with non idealities in memristor based neural networks |
| url | https://doi.org/10.1038/s41467-020-18098-0 |
| work_keys_str_mv | AT djoksas committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT pfreitas committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT zchai committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT whng committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT mbuckwell committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT cli committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT wdzhang committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT qxia committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT ajkenyon committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks AT amehonic committeemachinesauniversalmethodtodealwithnonidealitiesinmemristorbasedneuralnetworks |