公开(公告)号：US11461659B2

公开(公告)日：2022-10-04

申请号：US15871460

申请日：2018-01-15

Applicant: Huawei Technologies Co., Ltd.

Inventor： Dandan Tu ,  Jiajin Zhang

IPC: G06N3/08 ,  G06K9/62 ,  G06N3/04 ,  G06V10/44 ,  G06V40/16 ,  G06F17/16

Abstract: A feature set determining method includes obtaining, according to a received feature set determining request, data used for feature learning. The feature set determining request includes a learning objective of the feature learning. The method includes performing type analysis on the data to divide the data into first-type data and second-type data. The method includes performing semi-supervised learning on the first-type data to extract multiple first-type features. The method includes performing adaptive learning on the second-type data to extract multiple second-type features. The method includes evaluating the first-type features and the second-type features to obtain an optimal feature set.

公开(公告)号：US11003533B2

公开(公告)日：2021-05-11

申请号：US16369102

申请日：2019-03-29

Applicant: HUAWEI TECHNOLOGIES CO., LTD.

Inventor： Jiajin Zhang ,  Matt M. T. Yiu ,  Pak-Ching Lee

IPC: G11C29/00 ,  G06F11/10 ,  G06F3/06 ,  G06F12/02

Abstract: A data processing method is disclosed, and the method includes: encoding a data chunk of a predetermined size, to generate an error-correcting data chunk corresponding to the data chunk, where the data chunk includes a data object, and the data object includes a key, a value, and metadata; and generating a data chunk index and a data object index, where the data chunk index is used to retrieve the data chunk and the error-correcting data chunk corresponding to the data chunk, the data object index is used to retrieve the data object in the data chunk, and each data object index is used to retrieve a unique data object.

公开(公告)号：US20180150746A1

公开(公告)日：2018-05-31

申请号：US15871460

申请日：2018-01-15

Applicant: Huawei Technologies Co., Ltd.

Inventor： Dandan Tu ,  Jiajin Zhang

IPC: G06N3/08 ,  G06K9/62 ,  G06F17/16

CPC classification number: G06N3/088 ,  G06F17/16 ,  G06K9/00221 ,  G06K9/4628 ,  G06K9/6234 ,  G06K9/6271 ,  G06N3/0454

Abstract: A feature set determining method includes obtaining, according to a received feature set determining request, data used for feature learning. The feature set determining request includes a learning objective of the feature learning. The method includes performing type analysis on the data to divide the data into first-type data and second-type data. The method includes performing semi-supervised learning on the first-type data to extract multiple first-type features. The method includes performing adaptive learning on the second-type data to extract multiple second-type features. The method includes evaluating the first-type features and the second-type features to obtain an optimal feature set.