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.

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.

Abstract: Implementations of the present specification provide a value determination for a loss function. In an implementation, a first share of data that was generated based on a first parameter is received from a partner. A first term is determined based on the first share of the data. A second share of the data is generated based on a second parameter and a random number and transmitted to the partner. A coefficient of a second term is determined based on the random number. A second term is determined based on the coefficient of the second term and the second share of the data. The value of the loss function is determined based on the first term and the second term.

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.

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.

Abstract: This disclosure relates to secret sharing data exchange for generating a data processing model. In some aspects, first data party device determines respective values of first coefficients based on a first share of service data. The first coefficients are corresponding coefficients of respective target variables in different terms of a polynomial expression and the target variables are variables that are in the polynomial expression and associated with the first share of the service data. A second data party device determines respective values of second coefficients based on a second share of the service data. The second coefficients include coefficients other than the first coefficients in the different terms of the polynomial expression. The first data party device secretly shares respective values of the different terms in the polynomial expression in parallel based on the respective values of the first coefficients.

Abstract: This disclosure relates to secret sharing data exchange for generating a data processing model. In some aspects, first data party device determines respective values of first coefficients based on a first share of service data. The first coefficients are corresponding coefficients of respective target variables in different terms of a polynomial expression and the target variables are variables that are in the polynomial expression and associated with the first share of the service data. A second data party device determines respective values of second coefficients based on a second share of the service data. The second coefficients include coefficients other than the first coefficients in the different terms of the polynomial expression. The first data party device secretly shares respective values of the different terms in the polynomial expression in parallel based on the respective values of the first coefficients.

Abstract: Implementations of the present specification provide method for determining a new model parameter for a data processing model. The method includes determining and receiving, from one or more other data parties, one or more other shares of a new model parameter. The one or more other shares of the new model parameter can be combined with a first share of the new model parameter determined by a first data party to produce a new model parameter. The new model parameter is used to train the data processing model.

Abstract: Implementations of the present specification provide method for determining a new model parameter for a data processing model. The method includes determining and receiving, from one or more other data parties, one or more other shares of a new model parameter. The one or more other shares of the new model parameter can be combined with a first share of the new model parameter determined by a first data party to produce a new model parameter. The new model parameter is used to train the data processing model.

Abstract: Implementations this specification provide a method and an apparatus for determining data processing model parameters through multiparty cooperation. An example method performed by a data party device includes secretly sharing a first product with a cooperation partner device, based on characteristic data and a share of an original model parameter; communicating with the cooperation partner device, based on the share of the first product and a garbled circuit corresponding to an activation function, to obtain a share of a value of the activation function; secretly sharing a gradient of a loss function with the cooperation partner device, based on the characteristic data and the share of the value of the activation function, to obtain a share of the gradient of the loss function; and computing a share of a new model parameter.

Abstract: Implementations of the present specification provide a value determination for a loss function. In an implementation, a first share of data that was generated based on a first parameter is received from a partner. A first term is determined based on the first share of the data. A second share of the data is generated based on a second parameter and a random number and transmitted to the partner. A coefficient of a second term is determined based on the random number. A second term is determined based on the coefficient of the second term and the second share of the data. The value of the loss function is determined based on the first term and the second term.