# US 11,113,613 B2 · Determining model parameters using secret sharing > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,113,613 B2 - Document ID: US11113613B2 - Status: Granted - Kind: B2 - Application number: 17/188,837 - Filed: 2021-03-01 - Published: 2021-06-17 - Granted: 2021-09-07 - Earliest claimed: 2019-08-09 - Assignees: Advanced New Technologies Co., Ltd. - Applicants: Advanced New Technologies Co., Ltd. - Inventors: Yashun Zhou; Lichun Li; Shan Yin; Huazhong Wang - Web page: https://patentgrep.com/patent/US11113613B2 ## CPC - G06F 21/6245: {Protecting personal data, e.g. for financial or medical purposes} - G06N 5/04: Inference or reasoning models - G06F 17/16: Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)} - G06F 21/60: Protecting data - G06F 21/606: {by securing the transmission between two devices or processes} - G06N 20/00: Machine learning - H04L 9/085: {Secret sharing or secret splitting, e.g. threshold schemes} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract This disclosure relates to determining model parameters using secret sharing. In some aspects, a first data party device obtains a first share of a Hessian matrix for a data processing model. The first data party device obtains, using secret sharing with the second data party device, a first share of a product of a random number matrix and the Hessian matrix. The first data party device, determines a first share of a first inverse matrix based on a second inverse matrix and the first share of the random number matrix. The first data party device determines the first inverse matrix, a first share of a product of the first inverse matrix and a gradient of a loss function of the data processing model, and a first share of a new model parameter for the data processing model. ## Claims ### Claim 1 A computer-implemented method, comprising: obtaining, by a first data party device, a first share of a Hessian matrix for a data processing model, wherein the Hessian matrix is secretly shared between the first data party device and a second data party device of a cooperation partner such that the first data party device and the second data party device have respective shares of the Hessian matrix, and wherein the Hessian matrix is based on feature data for the data processing model and an activation function for the data processing model; obtaining, by the first data party device and using secret sharing with the second data party device, a first share of a product of a random number matrix and the Hessian matrix; determining, by the first data party device and using secret sharing with the second data party device, a first share of a product of a first inverse matrix and a gradient of a loss function of the data processing model, wherein the first inverse matrix is an inverse of the Hessian matrix; and determining a first share of a new model parameter for the data processing model based on an original model parameter of the data processing model and the first share of the product of the first inverse matrix and the gradient of the loss function. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, further comprising updating the data processing model with the first share of the new model parameter and a second share of the new model parameter obtained from the second data party device. ### Claim 3 (depends on 2) The computer-implemented method of claim 2, wherein the second share of the new model parameter is based on the original model parameter of the data processing model and a second share of the product of the first inverse matrix and the gradient of the loss function. ### Claim 4 (depends on 1) The computer-implemented method of claim 1, wherein the data processing model is a machine learning model and the new model parameter comprises one of a weight or a bias of the machine learning model. ### Claim 5 (depends on 1) The computer-implemented method of claim 1, further comprising: providing, to the second data party device, the first share of the product of the random number matrix and the Hessian matrix; in response to providing the first share of the product of the random number matrix and the Hessian matrix to the second data party device, receiving, by the first data party device, a second inverse matrix from the second data party device, wherein the second inverse matrix is an inverse of the product of the random number matrix and the Hessian matrix; and determining, by the first data party device, the first share of the first inverse matrix based on the second inverse matrix and a first share of the random number matrix. ### Claim 6 (depends on 5) The computer-implemented method of claim 5, further comprising determining, by the first data party device, the first inverse matrix based on the first share of the first inverse matrix and a second share of the first inverse matrix received from the second data party device, wherein the second share of the first inverse matrix is based on the second inverse matrix and a second share of the random number matrix. ### Claim 7 (depends on 1) The computer-implemented method of claim 1, wherein determining the first share of the new model parameter comprises subtracting the first share of the product of the first inverse matrix and the gradient of the loss function from the original model parameter. ### Claim 8 (depends on 7) The computer-implemented method of claim 7, wherein a second share of the product of the first inverse matrix and the gradient of the loss function comprises a product of the first inverse matrix and a second share of the gradient of the loss function secretly shared with the second data party device. ### Claim 9 A non-transitory, computer-readable medium storing one or more instructions executable by a computer system of a first data party device to perform operations comprising: obtaining a first share of a Hessian matrix for a data processing model, wherein the Hessian matrix is secretly shared between the first data party device and a second data party device of a cooperation partner such that the first data party device and the second data party device have respective shares of the Hessian matrix, and wherein the Hessian matrix is based on feature data for the data processing model and an activation function for the data processing model; obtaining, using secret sharing with the second data party device, a first share of a product of a random number matrix and the Hessian matrix; determining, using secret sharing with the second data party device, a first share of a product of a first inverse matrix and a gradient of a loss function of the data processing model, wherein the first inverse matrix is an inverse of the Hessian matrix; and determining a first share of a new model parameter for the data processing model based on an original model parameter of the data processing model and the first share of the product of and the gradient of the loss function. ### Claim 10 (depends on 9) The non-transitory, computer-readable medium of claim 9, wherein the operations comprise updating the data processing model with the first share of the new model parameter and a second share of the new model parameter obtained from the second data party device. ### Claim 11 (depends on 9) The non-transitory, computer-readable medium of claim 9, wherein the data processing model is a machine learning model and the new model parameter comprises one of a weight or a bias of the machine learning model. ### Claim 12 (depends on 9) The non-transitory, computer-readable medium of claim 9, wherein the operations comprise: providing, to the second data party device, the first share of the product of the random number matrix and the Hessian matrix; in response to providing the first share of the product of the random number matrix and the Hessian matrix to the second data party device, receiving a second inverse matrix from the second data party device, wherein the second inverse matrix is an inverse of the product of the random number matrix and the Hessian matrix; and determining the first share of the first inverse matrix based on the second inverse matrix and a first share of the random number matrix. ### Claim 13 (depends on 12) The non-transitory, computer-readable medium of claim 12, wherein the operations comprise determining the first inverse matrix based on the first share of the first inverse matrix and a second share of the first inverse matrix received from the second data party device, wherein the second share of the first inverse matrix is based on the second inverse matrix and a second share of the random number matrix. ### Claim 14 (depends on 9) The non-transitory, computer-readable medium of claim 9, wherein determining the first share of the new model parameter comprises subtracting the first share of the product of the first inverse matrix and the gradient of the loss function from the original model parameter. ### Claim 15 A computer-implemented system, comprising: one or more computers of a first data party device; and one or more computer memory devices interoperably coupled with the one or more computers and having tangible, non-transitory, machine-readable media storing one or more instructions that, when executed by the one or more computers, perform operations comprising: obtaining a first share of a Hessian matrix for a data processing model, wherein the Hessian matrix is secretly shared between the first data party device and a second data party device of a cooperation partner such that the first data party device and the second data party device have respective shares of the Hessian matrix, and wherein the Hessian matrix is based on feature data for the data processing model and an activation function for the data processing model; obtaining, using secret sharing with the second data party device, a first share of a product of a random number matrix and the Hessian matrix; determining, using secret sharing with the second data party device, a first share of a product of a first inverse matrix and a gradient of a loss function of the data processing model, wherein the first inverse matrix is an inverse of the Hessian matrix; and determining a first share of a new model parameter for the data processing model based on an original model parameter of the data processing model and the first share of the product of the first inverse matrix and the gradient of the loss function. ### Claim 16 (depends on 15) The computer-implemented system of claim 15, wherein the operations comprise updating the data processing model with the first share of the new model parameter and a second share of the new model parameter obtained from the second data party device. ### Claim 17 (depends on 16) The computer-implemented system of claim 16, wherein the second share of the new model parameter is based on the original model parameter of the data processing model and a second share of the product of the first inverse matrix and the gradient of the loss function. ### Claim 18 (depends on 15) The computer-implemented system of claim 15, wherein the data processing model is a machine learning model and the new model parameter comprises one of a weight or a bias of the machine learning model. ### Claim 19 (depends on 15) The computer-implemented system of claim 15, wherein the operations comprise: providing, to the second data party device, the first share of the product of the random number matrix and the Hessian matrix; in response to providing the first share of the product of the random number matrix and the Hessian matrix to the second data party device, receiving a second inverse matrix from the second data party device, wherein the second inverse matrix is an inverse of the product of the random number matrix and the Hessian matrix; and determining the first share of the first inverse matrix based on the second inverse matrix and a first share of the random number matrix. ### Claim 20 (depends on 19) The computer-implemented system of claim 19, wherein the operations comprise determining the first inverse matrix based on the first share of the first inverse matrix and a second share of the first inverse matrix received from the second data party device, wherein the second share of the first inverse matrix is based on the second inverse matrix and a second share of the random number matrix. ### Claim 21 (depends on 15) The computer-implemented system of claim 15, wherein determining the first share of the new model parameter comprises subtracting the first share of the product of the first inverse matrix and the gradient of the loss function from the original model parameter. ### Claim 22 (depends on 15) The computer-implemented system of claim 15, wherein a second share of the product of the first inverse matrix and the gradient of the loss function comprises a product of the first inverse matrix and a second share of the gradient of the loss function secretly shared with the second data party device.