Classification with a disordered dopantatom network in silicon

doi:10.1038/s41586-019-1901-0

GSTDTAP > 地球科学

DOI	10.1038/s41586-019-1901-0
	Classification with a disordered dopantatom network in silicon
	Vagnozzi, Ronald J.1; Maillet, Marjorie 1; Sargent, Michelle A.1; Khalil, Hadi 1; Johansen, Anne Katrine Z.1; Schwanekamp, Jennifer A.2; York, Allen J.1; Huang, Vincent 1; Nahrendorf, Matthias 3; Sadayappan, Sakthivel 2; Molkentin, Jeffery D.1,4
	2020-01-16
发表期刊	NATURE
ISSN	0028-0836
EISSN	1476-4687
出版年	2020
卷号	577 期号:7790 页码:341-+
文章类型	Article
语种	英语
国家	Netherlands
英文关键词	Classification is an important task at which both biological and artificial neural networks excel(1,2). In machine learning, nonlinear projection into a high-dimensional feature space can make data linearly separable(3,4), simplifying the classification of complex features. Such nonlinear projections are computationally expensive in conventional computers. A promising approach is to exploit physical materials systems that perform this nonlinear projection intrinsically, because of their high computational density(5), inherent parallelism and energy efficiency(6,7). However, existing approaches either rely on the systems' time dynamics, which requires sequential data processing and therefore hinders parallel computation(5,6,8), or employ large materials systems that are difficult to scale up(7). Here we use a parallel, nanoscale approach inspired by filters in the brain(1) and artificial neural networks(2) to perform nonlinear classification and feature extraction. We exploit the nonlinearity of hopping conduction(9-11) through an electrically tunable network of boron dopant atoms in silicon, reconfiguring the network through artificial evolution to realize different computational functions. We first solve the canonical two-input binary classification problem, realizing all Boolean logic gates(12) up to room temperature, demonstrating nonlinear classification with the nanomaterial system. We then evolve our dopant network to realize feature filters(2) that can perform four-input binary classification on the Modified National Institute of Standards and Technology handwritten digit database. Implementation of our material-based filters substantially improves the classification accuracy over that of a linear classifier directly applied to the original data(13). Our results establish a paradigm of silicon-based electronics for smallfootprint and energy-efficient computation(14).
领域	地球科学 ; 气候变化 ; 资源环境
收录类别	SCI-E
WOS记录号	WOS:000509570100027
WOS关键词	ACCELERATOR ; EVOLUTION
WOS类目	Multidisciplinary Sciences
WOS研究方向	Science & Technology - Other Topics
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.173/C666/handle/2XK7JSWQ/281244
专题	地球科学资源环境科学气候变化
作者单位	1.Univ Cincinnati, Dept Pediat, Cincinnati Childrens Hosp Med Ctr, Cincinnati, OH 45221 USA; 2.Univ Cincinnati, Dept Internal Med, Heart Lung & Vasc Inst, Cincinnati, OH USA; 3.Harvard Med Sch, Massachusetts Gen Hosp, Cardiovasc Res Ctr, Ctr Syst Biol,Dept Imaging, Boston, MA 02115 USA; 4.Cincinnati Childrens Hosp Med Ctr, Howard Hughes Med Inst, Cincinnati, OH 45229 USA
推荐引用方式 GB/T 7714	Vagnozzi, Ronald J.,Maillet, Marjorie,Sargent, Michelle A.,et al. Classification with a disordered dopantatom network in silicon[J]. NATURE,2020,577(7790):341-+.
APA	Vagnozzi, Ronald J..,Maillet, Marjorie.,Sargent, Michelle A..,Khalil, Hadi.,Johansen, Anne Katrine Z..,...&Molkentin, Jeffery D..(2020).Classification with a disordered dopantatom network in silicon.NATURE,577(7790),341-+.
MLA	Vagnozzi, Ronald J.,et al."Classification with a disordered dopantatom network in silicon".NATURE 577.7790(2020):341-+.