公开(公告)号：US20230410034A1

公开(公告)日：2023-12-21

申请号：US17845778

申请日：2022-06-21

Applicant: Google LLC

Inventor： Jiayu Peng ,  Chenwei Wang ,  Matthew Tran Clegg ,  Meng-Hao Li

IPC: G06Q10/08 ,  G06N7/00

CPC classification number: G06Q10/087 ,  G06N7/005

Abstract: In one aspect, there is provided a method that includes: obtaining multiple input frequency histograms that each correspond to a respective transmission commitment, where a transmission commitment corresponds to a subset of publishers from a set of publishers; generating a frequency model based on the input frequency histograms, where the frequency model is a parametric model parameterized by a set of model parameters that include a correlation matrix with a respective correlation value for each pair of publishers from the set of publishers; receiving a request to predict a frequency histogram for a target transmission commitment corresponding to a target subset of publishers; generating a predicted frequency histogram for the target transmission commitment using the frequency model; and generating one or more predictions characterizing the target transmission commitment using the predicted frequency histogram.

公开(公告)号：US20250005179A1

公开(公告)日：2025-01-02

申请号：US18502816

申请日：2023-11-06

Applicant: Google LLC

Inventor： Jiayu Peng ,  Michael James Wurm ,  Chenwei Wang ,  Pasin Manurangsi ,  Adam Benjamin Gelernter Sealfon ,  Jakub Tetek ,  Matthew Tran Clegg

IPC: G06F21/62

Abstract: Systems and methods for generating and maintaining differential privacy while providing accurate values can include obtaining a plurality of noise-added values, processing the plurality of noise-added values to determine a predicted value. The plurality of noise-added value may be utilized to determine one or more accuracy values that can be compared to a threshold to determine if more data is to be obtained and processed before providing a predicted value.

公开(公告)号：US20240005040A1

公开(公告)日：2024-01-04

申请号：US17856084

申请日：2022-07-01

Applicant: Google LLC

Inventor： Jiayu Peng ,  Meng-Hao Li ,  Chenwei Wang ,  Sanjay Lal Vasandani ,  Raimundo Mirisola

IPC: G06F21/62 ,  G06F16/2455

CPC classification number: G06F21/6263 ,  G06F16/24558 ,  G06F21/6227

Abstract: In one aspect, there is provided a method performed by one or more computers for privacy-sensitive assessment of digital component transmission reach based on cardinalities of subset unions of a collection of user sets, the method including: receiving a request to determine a number of users that are included in a target group of users that received at least one transmission of a digital component, where the request includes a set expression defined in terms of the collection of user sets, generating an alternative representation of the set expression in terms of primitive sets, applying a cardinality model to each primitive to generate a cardinality of each primitive set as a linear combination of cardinalities of subset unions of the collection of user sets, and determining the number of users included in the target group of users based on the cardinalities of the primitive sets.