Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators
Recently, a system of spintronic vortex oscillators has been experimentally trained to classify vowel sounds. In this paper, we have carried out a combination of device-level and system-level simulations to train a system of spin Hall nano oscillators (SHNOs) of smaller size (25X lower in area compa...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIP Publishing LLC
2021-04-01
|
| Series: | AIP Advances |
| Online Access: | http://dx.doi.org/10.1063/9.0000192 |
| _version_ | 1818595289181716480 |
|---|---|
| author | Utkarsh Singh Neha Garg Saurabh Kumar Pranaba Kishor Muduli Debanjan Bhowmik |
| author_facet | Utkarsh Singh Neha Garg Saurabh Kumar Pranaba Kishor Muduli Debanjan Bhowmik |
| author_sort | Utkarsh Singh |
| collection | DOAJ |
| description | Recently, a system of spintronic vortex oscillators has been experimentally trained to classify vowel sounds. In this paper, we have carried out a combination of device-level and system-level simulations to train a system of spin Hall nano oscillators (SHNOs) of smaller size (25X lower in area compared to those vortex oscillators) for such data classification tasks. Magnetic moments precess in an uniform mode as opposed to the vortex mode in our oscillators. We have trained our system to classify inputs in various popular machine learning data sets like Fisher’s Iris data set of flowers, Wisconsin Breast Cancer (WBC) data set, and MNIST data set of handwritten digits. We have employed a new technique for input dimensionality reduction here so that the clustering/target synchronization pattern changes based on the nature of the data in the different data sets. Our demonstration of learning in a system of such small SHNOs for a wide range of data sets is promising for scaling up the oscillator-based neuromorphic system for complex data classification tasks. |
| first_indexed | 2024-12-16T11:13:39Z |
| format | Article |
| id | doaj.art-022ba85a67704d089dbc1764d6a74d9e |
| institution | Directory Open Access Journal |
| issn | 2158-3226 |
| language | English |
| last_indexed | 2024-12-16T11:13:39Z |
| publishDate | 2021-04-01 |
| publisher | AIP Publishing LLC |
| record_format | Article |
| series | AIP Advances |
| spelling | doaj.art-022ba85a67704d089dbc1764d6a74d9e2022-12-21T22:33:39ZengAIP Publishing LLCAIP Advances2158-32262021-04-01114045117045117-1010.1063/9.0000192Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillatorsUtkarsh Singh0Neha Garg1Saurabh Kumar2Pranaba Kishor Muduli3Debanjan Bhowmik4Department of Electronics and Communication Engineering, Delhi Technological University, Delhi 110042, IndiaDepartment of Physics, Indian Institute of Technology Delhi, New Delhi 110016, IndiaDepartment of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi 110016, IndiaDepartment of Physics, Indian Institute of Technology Delhi, New Delhi 110016, IndiaDepartment of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi 110016, IndiaRecently, a system of spintronic vortex oscillators has been experimentally trained to classify vowel sounds. In this paper, we have carried out a combination of device-level and system-level simulations to train a system of spin Hall nano oscillators (SHNOs) of smaller size (25X lower in area compared to those vortex oscillators) for such data classification tasks. Magnetic moments precess in an uniform mode as opposed to the vortex mode in our oscillators. We have trained our system to classify inputs in various popular machine learning data sets like Fisher’s Iris data set of flowers, Wisconsin Breast Cancer (WBC) data set, and MNIST data set of handwritten digits. We have employed a new technique for input dimensionality reduction here so that the clustering/target synchronization pattern changes based on the nature of the data in the different data sets. Our demonstration of learning in a system of such small SHNOs for a wide range of data sets is promising for scaling up the oscillator-based neuromorphic system for complex data classification tasks.http://dx.doi.org/10.1063/9.0000192 |
| spellingShingle | Utkarsh Singh Neha Garg Saurabh Kumar Pranaba Kishor Muduli Debanjan Bhowmik Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators AIP Advances |
| title | Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators |
| title_full | Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators |
| title_fullStr | Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators |
| title_full_unstemmed | Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators |
| title_short | Learning of classification tasks with an array of uniform-mode spin Hall nano-oscillators |
| title_sort | learning of classification tasks with an array of uniform mode spin hall nano oscillators |
| url | http://dx.doi.org/10.1063/9.0000192 |
| work_keys_str_mv | AT utkarshsingh learningofclassificationtaskswithanarrayofuniformmodespinhallnanooscillators AT nehagarg learningofclassificationtaskswithanarrayofuniformmodespinhallnanooscillators AT saurabhkumar learningofclassificationtaskswithanarrayofuniformmodespinhallnanooscillators AT pranabakishormuduli learningofclassificationtaskswithanarrayofuniformmodespinhallnanooscillators AT debanjanbhowmik learningofclassificationtaskswithanarrayofuniformmodespinhallnanooscillators |