Title:
学習処理プログラム、情報処理装置及び学習処理方法
Document Type and Number:
Japanese Patent JP7368776
Kind Code:
B2
Abstract:
A deviation degree of a feature in a training dataset for each piece of data included in the training dataset is calculated by using a determination model that is generated through training by using the training dataset to which a label is not added, one or more pieces of data included in the training dataset or related data related to the one or more pieces of data is selected and output on the basis of the deviation degree, an input of a determination result by a user for the one or more pieces of data is received, an adjustment standard used to adjust the feature of each piece of the data included in the training dataset is determined on the basis of the received determination result, and a feature of determination target data is adjusted on the basis of the adjustment standard when the determination model determines the determination target data.
Inventors:
Ken Kobayashi
Yuhei Umeda
Yuhei Umeda
Application Number:
JP2022504775A
Publication Date:
October 25, 2023
Filing Date:
March 02, 2020
Export Citation:
Assignee:
富士通株式会社
International Classes:
G06N20/00
Other References:
Tiago Pimentel et al.,A Generalized Active Learning Approach for Unsupervised Anomaly Detection,arxiv.org,2018年05月23日,https://arxiv.org/abs/1805.09411v1
Attorney, Agent or Firm:
Kenji Doi
Tsunenori Hayashi
Tsunenori Hayashi
Previous Patent: Redundant operation system, redundant operation method, and program
Next Patent: Mold
Next Patent: Mold