[go: up one dir, main page]

WO2010035690A2 - Dispositif, système et procédé d'entraînement de structure, programme et support d'enregistrement - Google Patents

Dispositif, système et procédé d'entraînement de structure, programme et support d'enregistrement Download PDF

Info

Publication number
WO2010035690A2
WO2010035690A2 PCT/JP2009/066259 JP2009066259W WO2010035690A2 WO 2010035690 A2 WO2010035690 A2 WO 2010035690A2 JP 2009066259 W JP2009066259 W JP 2009066259W WO 2010035690 A2 WO2010035690 A2 WO 2010035690A2
Authority
WO
WIPO (PCT)
Prior art keywords
network structure
learning
network
agent
candidate
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/JP2009/066259
Other languages
English (en)
Japanese (ja)
Other versions
WO2010035690A9 (fr
Inventor
貴之 中田
森永　聡
遼平 藤巻
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NEC Corp
Original Assignee
NEC Corp
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 NEC Corp filed Critical NEC Corp
Priority to JP2010530824A priority Critical patent/JPWO2010035690A1/ja
Publication of WO2010035690A2 publication Critical patent/WO2010035690A2/fr
Publication of WO2010035690A9 publication Critical patent/WO2010035690A9/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the present invention relates to a structure learning device, a structure learning system, a structure learning method, a program, and a recording medium that improve the network structure learning result of the agent by sharing and using the network structure information of other agents.
  • Non-Patent Document 1 The learning method of the network structure learning system is described in Non-Patent Document 1, for example.
  • the network structure learning system described in Non-Patent Document 1 includes network structure learning means and network structure mixed parameter estimation means.
  • the network structure learning system having such a configuration operates as follows.
  • the network structure learning and the network structure mixed parameter estimation are performed by a method called an EM algorithm until the result converges.
  • Non-Patent Document 1 since the technique described in Non-Patent Document 1 does not consider sharing of the network structure by a plurality of agents, the plurality of agents share the network structure and cannot improve each learning result.
  • the present invention has been made in view of the above problems, and an object thereof is to provide a structure learning apparatus in which a plurality of agents can share a network structure and improve each learning result.
  • a structure learning device learns a network structure from a shared network and determines a candidate network structure, and a network structure made a candidate by the network structure learning means
  • Network structure mixed parameter estimating means for estimating the mixed parameters of the network structure at the network structure level
  • agent mixed parameter estimating means for estimating the network structure mixed parameters that are candidates by the network structure learning means at the agent level. It is characterized by.
  • the network structure generating means for generating a new network structure when the network structure learning means determines that there is no suitable network structure as a candidate.
  • the network structure learning means is characterized by repeatedly learning until the estimation results by the network structure mixed parameter estimating means and the agent mixed parameter estimating means converge.
  • the structure learning system according to the present invention includes a plurality of the structure learning devices described above.
  • the structure learning method is a network structure learning step for learning a network structure from a shared network and determining a network structure as a candidate, and a network structure mixed parameter determined by the network structure learning step as a network.
  • a network structure mixed parameter estimating step for estimating at a structure level, and an agent mixed parameter estimating step for estimating a network structure mixed parameter made a candidate by the network structure learning step at an agent level.
  • a network structure generation step is provided for generating a new network structure when it is determined in the network structure learning step that there is no network structure suitable as a candidate.
  • the network structure learning step is characterized in that learning is repeated until the estimation results of the network structure mixed parameter estimation step and the agent mixed parameter estimation step converge.
  • the program according to the present invention is characterized by causing a computer to execute the structure learning method described above.
  • the recording medium in the present invention is a computer-readable recording medium storing the above-described program.
  • a plurality of agents can share the network structure and improve each learning result.
  • FIG. 1 is a hardware configuration diagram of a network structure learning apparatus according to an embodiment of the present invention.
  • the apparatus includes an operation unit 1, a CPU 2, a memory 3, a ROM 4, and a network interface 5.
  • the CPU 2 controls the entire apparatus, and the ROM 4 stores programs executed by the CPU 2 and other fixed data.
  • the memory 3 is used for temporary storage of data when the CPU 2 performs calculations. Further, a network is realized by the network interface 5, and information can be input / output by the operation unit 1.
  • FIG. 2 is a system block diagram of the network structure learning apparatus according to the embodiment of the present invention.
  • the network structure learning apparatus includes an input unit 11, a network structure learning unit 12, a network structure generation unit 13, a network structure mixed parameter estimation unit 14, an agent mixed parameter estimation unit 15, and an output unit 16. Configured.
  • the input unit 11 takes in the data to be observed, and the network structure learning unit 12 learns the network structure that generates the observation data.
  • the network structure generation unit 13 generates a new network structure when the existing network structure does not match the observation value well by the network structure learning unit 12.
  • the network structure mixed parameter estimation means 14 estimates the mixing probability at the network structure level
  • the agent mixing parameter estimation means 15 estimates the mixing probability at the agent level.
  • the output means 16 outputs the learned network structure and the estimation result in the mixing probability.
  • step S1 data to be observed is captured by the input means 11 (step S1).
  • step S2 the network structure learning unit 12 learns the most suitable structure as a network structure for generating observation data (step S2). It is determined whether or not there is (step S3).
  • step S4 If there is no network structure that has been successfully matched by learning so far, a new most suitable network structure is generated from the distribution serving as the base of the network structure (step S4).
  • the network structure level is calculated to calculate what mixing probability the network structure has, and the mixing probability is estimated (step S5). It is determined whether or not it has converged (step S6). Here, when it does not converge, it returns to step S2 and learns again.
  • step S7 calculates what mixing probability the agent has at the agent level, estimates the mixing probability (step S7), and determines whether the estimation result has converged (step S8). .
  • step S8 determines whether the estimation result has converged.
  • step S9 output the network structure learned in step S2 and the mixing probability estimated in steps S5 and S7 (step S9).
  • the probability distribution T of the network is
  • Nj represents the number of observations of data at agent j.
  • the observation value is assumed to be exchangeable in the agent, and is generated according to the probability distribution xji to T ( ⁇
  • the network structure and its parameter ⁇ T: ji are generated from Gj,
  • nDP nested Dirichlet process
  • the index cj represents the mixed network structure that generates the network structure of agent j.
  • the index sji represents the network structure for generating the data point xji.
  • nDP and shared network structure can be performed based on Gibbs sampler, and the following steps are repeated until convergence.
  • (ii) and (iv) correspond to the network structure mixed parameter estimating means 14 of FIG. 2 and step S5 of FIG. 3, and (i) and (iii) are the agent mixed parameters of FIG. 3 corresponds to the estimation means 15 and step S7 in FIG. 3, and (v) corresponds to the network structure learning means 12, the network structure generation means 13 in FIG. 2, and steps S2 to S4 in FIG.
  • a plurality of agents can share a network structure and improve each learning result.
  • a mixed model is used for each agent model, a plurality of agents can be clustered into similar ones.
  • modeling, learning, and estimation are further performed using a nonparametric model, so that it is not necessary to set a finite value as the number of models mixed in advance.
  • the present invention can be used as follows.
  • the basic part correlation structure such as engine control shows similar behavior, but for example in the case of a hybrid vehicle, electronic motor control, etc. It has a parts correlation structure unique to the model.
  • the learning accuracy is improved by sharing information on other vehicle types, and the part correlation structure that can be found only in a specific vehicle type is learned independently for each vehicle type.
  • the present invention can be applied to such a case that vehicle types that share a part correlation structure with high uniqueness in the vehicle type are clustered as similar vehicle types.
  • the learning accuracy can be improved by sharing information of other banks.
  • each bank the number of specific operations, regional characteristics such as the occurrence of earthquakes and typhoons
  • the present invention can also be applied to applications such as sharing information between banks having the specific loss event correlation structure and clustering them as banks having a similar loss event correlation structure.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Artificial Intelligence (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

L'invention concerne un dispositif d'entraînement de structure dans lequel une pluralité d'agents partagent une structure de réseau, et le résultat d'entraînement de chaque agent peut être amélioré. Le dispositif comprend : un moyen d'entraînement de structure de réseau destiné à enseigner une structure de réseau d'un réseau partagé et à déterminer une structure de réseau candidate, un moyen d'évaluation de paramètres mélangés au niveau structure de réseau destiné à évaluer les paramètres mélangés de la structure de réseau candidate du moyen d'entraînement de structure de réseau au niveau de la structure de réseau, et un moyen d'évaluation de paramètres mélangés au niveau agent destiné à évaluer des paramètres mélangés de la structure de réseau candidate du moyen d'entraînement de structures de réseau au niveau agent.
PCT/JP2009/066259 2008-09-24 2009-09-17 Dispositif, système et procédé d'entraînement de structure, programme et support d'enregistrement Ceased WO2010035690A2 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2010530824A JPWO2010035690A1 (ja) 2008-09-24 2009-09-17 構造学習装置、構造学習システム、構造学習方法、プログラム及び記録媒体

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2008-244168 2008-09-24
JP2008244168 2008-09-24

Publications (2)

Publication Number Publication Date
WO2010035690A2 true WO2010035690A2 (fr) 2010-04-01
WO2010035690A9 WO2010035690A9 (fr) 2011-02-03

Family

ID=42060202

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2009/066259 Ceased WO2010035690A2 (fr) 2008-09-24 2009-09-17 Dispositif, système et procédé d'entraînement de structure, programme et support d'enregistrement

Country Status (2)

Country Link
JP (1) JPWO2010035690A1 (fr)
WO (1) WO2010035690A2 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111602148A (zh) * 2018-02-02 2020-08-28 谷歌有限责任公司 正则化神经网络架构搜索

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106156856A (zh) * 2015-03-31 2016-11-23 日本电气株式会社 用于混合模型选择的方法和装置

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111602148A (zh) * 2018-02-02 2020-08-28 谷歌有限责任公司 正则化神经网络架构搜索
US11669744B2 (en) 2018-02-02 2023-06-06 Google Llc Regularized neural network architecture search
CN111602148B (zh) * 2018-02-02 2024-04-02 谷歌有限责任公司 正则化神经网络架构搜索
US12400121B2 (en) 2018-02-02 2025-08-26 Google Llc Regularized neural network architecture search

Also Published As

Publication number Publication date
JPWO2010035690A1 (ja) 2012-02-23
WO2010035690A9 (fr) 2011-02-03

Similar Documents

Publication Publication Date Title
CN115943394B (zh) 用于安全纵向联邦学习的方法、装置和系统
Berg et al. Structure and evolution of protein interaction networks: a statistical model for link dynamics and gene duplications
Krull et al. Poseagent: Budget-constrained 6d object pose estimation via reinforcement learning
CN113344016B (zh) 深度迁移学习方法、装置、电子设备及存储介质
Dulac-Arnold et al. Fast reinforcement learning with large action sets using error-correcting output codes for mdp factorization
KR101700030B1 (ko) 사전 정보를 이용한 영상 물체 탐색 방법 및 이를 수행하는 장치
Liu et al. WTFM layer: An effective map extractor for unsupervised shape correspondence
Zhu et al. Trivialized momentum facilitates diffusion generative modeling on lie groups
Langford et al. Learning nonlinear dynamic models
Zeng et al. A nearsighted force-training approach to systematically generate training data for the machine learning of large atomic structures
Fan et al. Image steganalysis via random subspace fisher linear discriminant vector functional link network and feature mapping
Ruan et al. A robust and scalable solution for interpolative multidimensional scaling with weighting
Meier et al. Efficient Bayesian local model learning for control
WO2010035690A2 (fr) Dispositif, système et procédé d'entraînement de structure, programme et support d'enregistrement
CN116070437A (zh) 锂电池表面温度建模方法、装置、计算机设备和存储介质
Wong et al. State-space decomposition for reinforcement learning
Chen et al. Handwritten CAPTCHA recognizer: a text CAPTCHA breaking method based on style transfer network
Clementi et al. Folding Lennard‐Jones proteins by a contact potential
Chiu et al. Automatic complexity reduction in reinforcement learning
Varshney et al. Reduced order modeling and dynamic optimization of multiscale PDE/kMC process systems
JP2010134926A (ja) マルチ・モーダル・グリーン・コンピューティング統合システム、コンピュータ・プログラム、方法、およびシステムの展開方法（問題解析を用いたマルチ・モーダル・グリーン・コンピューティング統合）
CN119790405A (zh) 基于模型的任务处理
WO2020040007A1 (fr) Dispositif d'apprentissage, procédé d'apprentissage et programme d'apprentissage
CN114255374B (zh) 一种针对软间隔支持向量机的量子方法
Wang et al. Generalized flow matching for transition dynamics modeling

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 09816105

Country of ref document: EP

Kind code of ref document: A2

WWE Wipo information: entry into national phase

Ref document number: 2010530824

Country of ref document: JP

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 09816105

Country of ref document: EP

Kind code of ref document: A2