BR112013032604A2 - metodo e aparelho para regra de aprendizagem competitiva local que leva a conectividade esparsa - Google Patents

metodo e aparelho para regra de aprendizagem competitiva local que leva a conectividade esparsa

Info

Publication number
BR112013032604A2
BR112013032604A2 BR112013032604A BR112013032604A BR112013032604A2 BR 112013032604 A2 BR112013032604 A2 BR 112013032604A2 BR 112013032604 A BR112013032604 A BR 112013032604A BR 112013032604 A BR112013032604 A BR 112013032604A BR 112013032604 A2 BR112013032604 A2 BR 112013032604A2
Authority
BR
Brazil
Prior art keywords
rule
sparse
local
learning rule
oja
Prior art date
Application number
BR112013032604A
Other languages
English (en)
Inventor
Aparin Vladimir
Original Assignee
Qualcomm Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Qualcomm Inc filed Critical Qualcomm Inc
Publication of BR112013032604A2 publication Critical patent/BR112013032604A2/pt

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/0205—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system
    • G05B13/024—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system in which a parameter or coefficient is automatically adjusted to optimise the performance
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
    • G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/088—Non-supervised learning, e.g. competitive learning

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Automation & Control Theory (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Neurology (AREA)
  • Complex Calculations (AREA)
  • Image Analysis (AREA)
  • Medicines Containing Plant Substances (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Feedback Control In General (AREA)

Abstract

resumo “método e aparelho para regra de aprendizagem competitiva local que leva a conectividade esparsa”. determinados aspectos da presente invenção suportam uma regra de aprendizagem competitiva local aplicada em uma rede computacional que leva à conectividade esparsa entre as unidades de processamento da rede. a presente invenção provê uma modificação na regra de aprendizagem de oja, que modifica a restrição à soma de pesos elevados ao quadrado na regra de oja. esta restrição pode ser intrínseca e local em oposição às normalizações multiplicadora e subtrativa comumente utilizadas, que são explícitas e exigem o conhecimento de todos os pesos de entrada de uma unidade de processamento para a atualização de cada um deles individualmente. a regra apresentada provê convergência para um vetor de peso que é mais esparso (isto é, tem mais elementos zero) que o vetor de peso aprendido pela regra de oja original. tal conectividade esparsa pode levar a uma seletividade mais elevada das unidades de processador com relação a recursos específicos e pode exigir menos memória para armazenar a configuração de rede e menos energia para acioná-la. 1/1
BR112013032604A 2011-06-22 2012-06-21 metodo e aparelho para regra de aprendizagem competitiva local que leva a conectividade esparsa BR112013032604A2 (pt)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US13/166,269 US9129222B2 (en) 2011-06-22 2011-06-22 Method and apparatus for a local competitive learning rule that leads to sparse connectivity
PCT/US2012/043594 WO2012177913A1 (en) 2011-06-22 2012-06-21 Method and apparatus for a local competitive learning rule that leads to sparse connectivity

Publications (1)

Publication Number Publication Date
BR112013032604A2 true BR112013032604A2 (pt) 2017-08-01

Family

ID=46384523

Family Applications (1)

Application Number Title Priority Date Filing Date
BR112013032604A BR112013032604A2 (pt) 2011-06-22 2012-06-21 metodo e aparelho para regra de aprendizagem competitiva local que leva a conectividade esparsa

Country Status (9)

Country Link
US (1) US9129222B2 (pt)
EP (1) EP2724297B1 (pt)
JP (1) JP5819522B2 (pt)
KR (1) KR101549796B1 (pt)
CN (1) CN103620624B (pt)
BR (1) BR112013032604A2 (pt)
CA (1) CA2839279C (pt)
RU (1) RU2586864C2 (pt)
WO (1) WO2012177913A1 (pt)

Families Citing this family (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8583577B2 (en) * 2011-05-25 2013-11-12 Qualcomm Incorporated Method and apparatus for unsupervised training of input synapses of primary visual cortex simple cells and other neural circuits
US8843426B2 (en) * 2011-05-25 2014-09-23 Qualcomm Incorporated Method and apparatus of primary visual cortex simple cell training and operation
US11289175B1 (en) * 2012-11-30 2022-03-29 Hrl Laboratories, Llc Method of modeling functions of orientation and adaptation on visual cortex
JP5953414B1 (ja) * 2015-09-30 2016-07-20 美濃窯業株式会社 加熱炉の炉壁構造
CN109313720B (zh) * 2016-02-05 2023-07-18 渊慧科技有限公司 具有稀疏访问的外部存储器的增强神经网络
CN110298443B (zh) * 2016-09-29 2021-09-17 中科寒武纪科技股份有限公司 神经网络运算装置及方法
US11544545B2 (en) 2017-04-04 2023-01-03 Hailo Technologies Ltd. Structured activation based sparsity in an artificial neural network
US10387298B2 (en) 2017-04-04 2019-08-20 Hailo Technologies Ltd Artificial neural network incorporating emphasis and focus techniques
US11615297B2 (en) 2017-04-04 2023-03-28 Hailo Technologies Ltd. Structured weight based sparsity in an artificial neural network compiler
US11551028B2 (en) 2017-04-04 2023-01-10 Hailo Technologies Ltd. Structured weight based sparsity in an artificial neural network
US11238334B2 (en) 2017-04-04 2022-02-01 Hailo Technologies Ltd. System and method of input alignment for efficient vector operations in an artificial neural network
RU2702978C1 (ru) * 2018-10-15 2019-10-14 Самсунг Электроникс Ко., Лтд. Байесовское разреживание рекуррентных нейронных сетей
US11568237B2 (en) 2018-05-10 2023-01-31 Samsung Electronics Co., Ltd. Electronic apparatus for compressing recurrent neural network and method thereof
CN109740739B (zh) * 2018-12-29 2020-04-24 中科寒武纪科技股份有限公司 神经网络计算装置、神经网络计算方法及相关产品
US11461645B2 (en) * 2019-12-02 2022-10-04 International Business Machines Corporation Initialization of memory networks
US11263077B1 (en) 2020-09-29 2022-03-01 Hailo Technologies Ltd. Neural network intermediate results safety mechanism in an artificial neural network processor
US11221929B1 (en) 2020-09-29 2022-01-11 Hailo Technologies Ltd. Data stream fault detection mechanism in an artificial neural network processor
US11237894B1 (en) 2020-09-29 2022-02-01 Hailo Technologies Ltd. Layer control unit instruction addressing safety mechanism in an artificial neural network processor
US11874900B2 (en) 2020-09-29 2024-01-16 Hailo Technologies Ltd. Cluster interlayer safety mechanism in an artificial neural network processor
US11811421B2 (en) 2020-09-29 2023-11-07 Hailo Technologies Ltd. Weights safety mechanism in an artificial neural network processor

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5293456A (en) * 1991-06-28 1994-03-08 E. I. Du Pont De Nemours And Company Object recognition system employing a sparse comparison neural network
US5706400A (en) * 1995-03-08 1998-01-06 Nec Research Institute, Inc. Fault-tolerant implementation of finite-state automata in recurrent neural networks
JP2002024199A (ja) * 1998-02-20 2002-01-25 Souwa Kenkyusho:Kk 二値システムの学習方法
US6366897B1 (en) 1999-07-26 2002-04-02 Hnc Software, Inc. Cortronic neural networks with distributed processing
CA2402916A1 (en) 2000-03-16 2001-09-20 Yuan Yan Chen Apparatus and method for fuzzy analysis of statistical evidence
US7080053B2 (en) 2000-08-16 2006-07-18 Research Foundation Of State University Of New York Neural network device for evolving appropriate connections
US6954744B2 (en) * 2001-08-29 2005-10-11 Honeywell International, Inc. Combinatorial approach for supervised neural network learning
US20040034633A1 (en) * 2002-08-05 2004-02-19 Rickard John Terrell Data search system and method using mutual subsethood measures
JP4975287B2 (ja) 2005-08-08 2012-07-11 パナソニック株式会社 予測装置
US7962429B2 (en) 2007-05-24 2011-06-14 Paul Adams Neuromorphic device for proofreading connection adjustments in hardware artificial neural networks
CN100580698C (zh) * 2007-09-10 2010-01-13 东北大学 稀疏数据过程建模方法

Also Published As

Publication number Publication date
US20120330870A1 (en) 2012-12-27
CN103620624A (zh) 2014-03-05
EP2724297A1 (en) 2014-04-30
EP2724297B1 (en) 2018-09-05
JP2014520349A (ja) 2014-08-21
CA2839279C (en) 2017-10-03
WO2012177913A1 (en) 2012-12-27
CA2839279A1 (en) 2012-12-27
KR101549796B1 (ko) 2015-09-11
JP5819522B2 (ja) 2015-11-24
US9129222B2 (en) 2015-09-08
CN103620624B (zh) 2016-10-19
RU2014101717A (ru) 2015-07-27
KR20140027498A (ko) 2014-03-06
RU2586864C2 (ru) 2016-06-10

Similar Documents

Publication Publication Date Title
BR112013032604A2 (pt) metodo e aparelho para regra de aprendizagem competitiva local que leva a conectividade esparsa
BR112021016111A2 (pt) Dispositivo de computação, unidade de processamento paralelo, núcleo de unidade de processamento gráfico de propósito geral e multiprocessador gráfico
BR112019004798A8 (pt) Método implantado por computador e mídia de armazenamento
BR112014022638A8 (pt) Método, suporte físico e equipamento para transformar especificadores de instrução de um ambiente computacional
US20170116157A1 (en) Parallelizing matrix factorization across hardware accelerators
BRPI0819170A8 (pt) Método para executar uma tarefa computacional, e, sistema de computador em rede configurado para executar uma tarefa computacional
BR112018002040A2 (pt) controle de uma nuvem de dispositivo
BR112016006284A2 (pt) método para conduzir evolução direcionada, produto de programa de computador, e, sistema de computador
BR112016015618A8 (pt) decodificador, produto de programa de computador e dispositivo de hardware de rede para representação dos vetores de movimento em um fluxo contínuo de bits codificado
ES2709275T3 (es) Acceso multi-inquilino a aplicaciones
BR112015004684A2 (pt) geração de código nativo a partir de código de linguagem intermediária para um aplicativo
BR112018076877A8 (pt) Método para processamento de recursos e aparelho para processar um recurso
BR112019004964A2 (pt) gerador de caso de teste construído em editor de fluxo de trabalho de integração de dados
BR112013028501A2 (pt) aparelho e método para processamento de dados seguro baseado em hardware utilizando regras de faixa de endereço de memória de armazenamento temporário
BR112016024418A8 (pt) métodos implementados por computador e meios de armazenamento em computador para balanceamento de carga com base em dimensão
BR112014003298A2 (pt) artigo fabricado compreendendo um meio lido por computador, aparelho e método
BR112017010075A2 (pt) processadores, métodos, sistemas e instruções de ajuste de coordenada de morton
BR112014010237A2 (pt) processamento de dados de imagem e método
BR112014010543A2 (pt) sistemas e métodos para estabilização de oleoduto submarino
BR112017013350A2 (pt) atenuação térmica em aparelhos com várias unidades de processamento
BR112015023491A2 (pt) granularidade controlada por aplicativo para classificação eficaz em potência
BR112013028524A2 (pt) método para configurar pelo menos um dispositivo de processamento de entrada/saída configurável, dispositivo de processamento de entrada/saída configurável, sistema de controle de aviônica, programa de computador e produto de programa de computador
Barrios et al. D3GB: an interactive genome browser for R, Python, and WordPress
BR112018076270A2 (pt) método e aparelho para realizar operações de coleta e cópia simd
Holmes et al. Huddersfield university campus grid: Qgg of oscar clusters

Legal Events

Date Code Title Description
B06F Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]
B06U Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]
B07A Technical examination (opinion): publication of technical examination (opinion) [chapter 7.1 patent gazette]
B09B Patent application refused [chapter 9.2 patent gazette]
B12B Appeal: appeal against refusal