公开(公告)号：US20240005210A1

公开(公告)日：2024-01-04

申请号：US18252559

申请日：2021-11-06

Applicant: Lemon Inc.

Inventor： Jiankai SUN ,  Weihao GAO ,  Chong WANG ,  Hongyi ZHANG ,  Xiaobing LIU ,  Runliang LI ,  Xin YANG

IPC: G06N20/00 ,  G06F21/60

CPC classification number: G06N20/00 ,  G06F21/60

Abstract: The present disclosure relates to a data protection method, an apparatus, a medium and a device. The method includes: acquiring gradient association information respectively corresponding to reference samples of a target batch of an active party of a joint training model; according to the proportion occupied respectively by reference samples of positive examples and reference samples of negative examples in all reference samples of the target batch, determining a constraint condition of the data noise to be added; determining information of said data noise according to the gradient association information and the constraint condition corresponding to the reference samples; correcting, according to the information of said data noise, an initial gradient transmission value corresponding to each reference sample, so as to obtain target gradient transmission information; and sending the target gradient transmission information to a passive party of the joint training model.

公开(公告)号：US20240330707A1

公开(公告)日：2024-10-03

申请号：US18623714

申请日：2024-04-01

Applicant: Douyin Vision Co., Ltd. ,  Lemon Inc.

Inventor： Anran XU ,  Shengjie WANG ,  Lei CHEN ,  Shuai LI ,  Jiankai SUN ,  Zhenzhe ZHENG ,  Fan WU

IPC: G06N3/098

CPC classification number: G06N3/098

Abstract: A method, apparatus, device, and medium for action execution are provided. In a method, a set of actions to be executed at a first device is determined from a plurality of actions based on a first action model at the first device. A data accumulated indicator associated with the set of actions is obtained, the data accumulated indicator indicating an amount of data to be sent from the first device to a second device associated with the first device. In response to that the data accumulated indicator meets a predetermined condition, parameter data associated with the set of actions is transmitted to the second device to cause the second device to update a second action model at the second device using the parameter data, the parameter data comprising reward data and consumption data associated with the set of actions respectively.

公开(公告)号：US20240311647A1

公开(公告)日：2024-09-19

申请号：US18604023

申请日：2024-03-13

Applicant: Douyin Vision Co., Ltd. ,  Lemon Inc.

Inventor： Xinwei WAN ,  Jiankai SUN ,  Shengjie WANG ,  Lei CHEN ,  Zhenzhe ZHENG ,  Fan WU ,  Guihai CHEN

IPC: G06N3/098 ,  G06N3/045

CPC classification number: G06N3/098 ,  G06N3/045

Abstract: A method according to embodiments of the present disclosure includes generating a multi-classification label set corresponding to an object set based on a binary classification label set corresponding to the object set. The method further includes receiving an embedding vector set from a non-label party model, wherein an embedding vector in the embedding vector set is generated based on a feature of an object in the object set. The method further includes generating a label party model based on the embedding vector set and the multi-classification label set, wherein the label party model includes a first network and a second network. The method according to embodiments of the present disclosure enables a label party to protect privacy of an original label set under the condition of joint training with a non-label party, and prevent the non-label party from inferring original labels corresponding to original features by various means.

公开(公告)号：US20240242089A1

公开(公告)日：2024-07-18

申请号：US18565015

申请日：2022-04-28

Applicant: Lemon Inc.

Inventor： Jiankai SUN ,  Weihao GAO ,  Junyuan XIE ,  Chong WANG

IPC: G06N3/098 ,  G06N3/04

CPC classification number: G06N3/098 ,  G06N3/04

Abstract: The present disclosure relates to a data protection method, a training method and apparatus for a network structure, a medium, and a device. The data protection method includes: obtaining original feature information of a target batch of reference samples for a passive party of a joint training model; and processing the original feature information by means of a target feature processing network structure to obtain target feature information corresponding to the original feature information. A neural network structure is trained by at least aiming at minimizing a coupling degree of between original training feature information and target training feature information of training samples for the passive party to obtain the target feature processing network structure. The target training feature information is feature information corresponding to the original training feature information that is outputted from the neural network structure using the original training feature information as an input.

公开(公告)号：US20240249004A1

公开(公告)日：2024-07-25

申请号：US18565000

申请日：2022-04-28

Applicant: Lemon Inc.

Inventor： Xin YANG ,  Jiankai SUN ,  Weihao GAO ,  Junyuan XIE ,  Chong WANG

IPC: G06F21/60 ,  G06N20/20

CPC classification number: G06F21/602 ,  G06N20/20

Abstract: The present disclosure relates to a method and a device for data protection, a readable medium and an electronic apparatus, and the method comprises: acquiring a target identification information union set, wherein the target identification information union set comprises target encryption identification information of a first party of a joint training model and target encryption identification information of a second party of the joint training model, the target encryption identification information in the target identification information union set being obtained by encrypting according to a secret key of the first party and a secret key of the second party; and determining, according to the target identification information union set, a target sample data set for training the joint training model. Therefore, an identification information intersection of the first party and the second party does not need to be determined in advance as in the related technology.

公开(公告)号：US20240220641A1

公开(公告)日：2024-07-04

申请号：US18565962

申请日：2022-07-15

Applicant: Lemon Inc.

Inventor： Jiankai SUN ,  Xin YANG ,  Aonan ZHANG ,  Weihao GAO ,  Junyuan XIE ,  Chong WANG

IPC: G06F21/60 ,  G06N20/00

CPC classification number: G06F21/602 ,  G06N20/00

Abstract: The present disclosure relates to a data protection method, apparatus, medium and electronic device. The method comprises: obtaining a specified batch of reference samples of an active participant of a joint training model; determining generation gradient information of the first reference sample; determining target gradient information sent to the passive participant according to the generation gradient information, and sending the target gradient information to the passive participant, to update, by the passive participant, parameters of the joint training model according to the target gradient information. Through the above solution, the influence of the generated data on the training process and model performance of the joint training model is avoided as much as possible, and the privacy and security of data are improved.

公开(公告)号：US20240126899A1

公开(公告)日：2024-04-18

申请号：US18539851

申请日：2023-12-14

Applicant: Lemon Inc.

Inventor： Xin YANG ,  Junyuan XIE ,  Jiankai SUN ,  Yuanshun YAO ,  Chong WANG

IPC: G06F21/62 ,  G06N3/04 ,  G06N3/098

CPC classification number: G06F21/62 ,  G06N3/04 ,  G06N3/098

Abstract: There are proposed a method, device, apparatus, and medium for protecting sensitive data. In a method, to-be-processed data is received from a server device. A processing result of a user for the to-be-processed data is received, the processing result comprising sensitive data of the user for the processing of the to-be-processed data. A gradient for training a server model at the server device is determined based on a comparison between the processing result and a prediction result for the to-be-processed data. The gradient is updated in a change direction associated with the gradient so as to generate an updated gradient to be sent to the server device. Noise is added only in the change direction associated with the gradient. The corresponding overhead of processing noise in a plurality of directions can be reduced, and no excessive noise data interfering with training will be introduced to the updated gradient.

公开(公告)号：US20240119341A1

公开(公告)日：2024-04-11

申请号：US17953255

申请日：2022-09-26

Applicant: Lemon Inc.

Inventor： Xin YANG ,  Hanlin ZHU ,  Tianyi LIU ,  Jiankai SUN ,  Yuanshun YAO ,  Aonan ZHANG ,  Chong WANG

IPC: G06N20/00

CPC classification number: G06N20/00

Abstract: The present disclosure describes techniques for determining performance of a classifier. A first machine learning model and a second machine learning model may be trained by aggregating updates to the first machine learning model and the second machine learning model received from a plurality of client computing devices. A cumulative distribution function (CDF) associated with a distribution of the positive samples in the user data may be estimated using the trained first machine learning model. A probability density function (PDF) associated with a distribution of the negative samples in the user data may be estimated using the trained second machine learning model. An integration-based computation of an area under the receiver operating characteristic curve (AUC) of the classifier may be performed using the PDF and the CDF.