公开(公告)号：US20180181875A1

公开(公告)日：2018-06-28

申请号：US15128539

申请日：2015-03-11

Applicant: NEC Corporation

Inventor： Yousuke Motohashi ,  Masato Asahara ,  Satoshi Morinaga

IPC: G06N99/00 ,  G06N5/04 ,  G06Q30/02

CPC classification number: G06N20/00 ,  G06N5/04 ,  G06N5/045 ,  G06Q10/04 ,  G06Q30/0283

Abstract: In this invention, a property of a prediction target or analysis target can be predicted or analyzed with a high degree of precision during a transition from a stage in which there is extremely little or no known data about said prediction target or analysis target to a stage in which a sufficient amount of known data has been accumulated. This learning-model selection system comprises a model-evaluating means for evaluating learning models and a model-selecting means for selecting either a target learning model or a higher-order learning model on the basis of the result of the evaluation.

公开(公告)号：US11610151B2

公开(公告)日：2023-03-21

申请号：US16619829

申请日：2018-06-05

Applicant: NEC CORPORATION

Inventor： Masato Asahara ,  Ryohei Fujimaki ,  Yusuke Muraoka

IPC: G06F1/00 ,  G06N20/00 ,  G06N5/04 ,  G06N7/00

Abstract: A distribution system 100 includes a data management apparatus 10 and a plurality of calculators 20 that execute machine learning. The data management apparatus 10 includes a data acquisition unit 11 that acquires information regarding training data held in a memory 21 of each of the calculators 20, from the calculators 20, and a data rearrangement unit 12 that determines training data that is to be held in the memory 21 of each of the calculators 20, based on characteristics of the machine learning processes that are executed by the calculators 20, and the information acquired from the calculators.

公开(公告)号：US10255114B2

公开(公告)日：2019-04-09

申请号：US14900718

申请日：2014-03-06

Applicant: NEC Corporation

Inventor： Masato Asahara ,  Kazuyo Narita ,  Jianquan Liu

IPC: H04L12/24 ,  G06F9/50 ,  G06F11/34 ,  H04L12/26 ,  H04L29/08

Abstract: An abnormality detection apparatus (2000) handles tasks allocated to a plurality of processing servers (3200) as processing targets in a distribution system (3000) having the processing servers (3200). A history acquisition unit (2020) acquires progress history information which is information regarding progress of the plurality of tasks at a plurality of time point of recording. A target range determination unit (2040) determines a target range. A distribution calculation unit (2060) calculates a task speed distribution which is a probability distribution of processing speeds of the tasks using the progress history information regarding the plurality of tasks. An abnormality determination unit (2080) compares a processing speed of a task to be determined with the task speed distribution to thereby determine whether or not the processing speed of the task to be determined is abnormal.

公开(公告)号：US11656917B2

公开(公告)日：2023-05-23

申请号：US16759202

申请日：2018-10-25

Applicant: NEC CORPORATION

Inventor： Masato Asahara

IPC: G06E1/00 ,  G06F9/50 ,  G06N20/00 ,  G06F9/54

CPC classification number: G06F9/5083 ,  G06F9/541 ,  G06N20/00

Abstract: A distributed processing management apparatus 10 is connected to a plurality of execution servers 20 so as to be able to communicate with the execution servers 20. The distributed processing management apparatus 10 is provided with a conversion instruction unit 11 configured to specify, for each execution server 20, a data format usable by a machine learning engine executed by the execution server 20, and issue an instruction to convert a data format of data held by the execution server 20 to the specified data format.

公开(公告)号：US11204805B2

公开(公告)日：2021-12-21

申请号：US16608113

申请日：2018-04-09

Applicant: NEC CORPORATION

Inventor： Akihiro Yabe ,  Masato Asahara ,  Ryohei Fujimaki

IPC: G06F9/48 ,  G06F9/50

Abstract: A computational resource management apparatus is for managing a cluster system that executes a plurality of tasks. The computational resource management apparatus includes a condition specification unit that specifies a relationship between computational resources of the cluster system and computation time, a dependency relationship between tasks, and an execution time limit of each task, and a scheduling unit that determines, for each task, an execution sequence and computational resources to be allocated from among the computational resources of the cluster system, based on the relationship between the computational resources and computation time and the dependency relationship that are specified, such that the execution time limit is met.

公开(公告)号：US10963297B2

公开(公告)日：2021-03-30

申请号：US16085097

申请日：2017-04-27

Applicant: NEC Corporation

Inventor： Masato Asahara ,  Akihiro Yabe ,  Kyota Kanno ,  Ryohei Fujimaki

IPC: G06F9/48 ,  G06N20/00 ,  G06F11/34 ,  G06F9/50 ,  G06N5/00 ,  G06N7/00

Abstract: A computational resource management device uses a measured value of an execution time of data processing, a measured value of a resource amount, and a feature of input data as training data to learn a model indicating a relationship between the execution time and the resource. The device inputs, into the model, a feature of data scheduled to be input to calculate an estimated value of the execution time of the scheduled data processing, and uses the estimated value of the execution time, a variation index indicating variation in the estimated value of the execution time, and distribution of estimated residuals to calculate a resource amount required in the scheduled data processing. The device creates an execution plan of the scheduled data processing, based on the estimated value of the execution time, the variation index, the distribution of estimated residuals, and the calculated resource amount.