CN111598254A - Federated learning modeling method, device and readable storage medium - Google Patents

Abstract

The application discloses a federated learning modeling method, equipment and a readable storage medium, wherein the federated learning modeling method comprises the following steps: receiving encryption model parameters sent by each second device and verification parameters corresponding to the encryption model parameters, respectively performing zero knowledge verification on each encryption model parameter based on each verification parameter to determine false encryption model parameters in each encryption model parameter to obtain a zero knowledge verification result, and coordinating each second device to perform federated learning modeling based on the zero knowledge verification result and each encryption model parameter. The method and the device solve the technical problems of low efficiency and poor accuracy of federal learning modeling.

Description

Translated fromChinese

联邦学习建模方法、设备及可读存储介质Federated learning modeling method, device and readable storage medium

技术领域technical field

本申请涉及金融科技(Fintech)的人工智能领域，尤其涉及一种联邦学习建模方法、设备及可读存储介质。The present application relates to the field of artificial intelligence of financial technology (Fintech), and in particular, to a federated learning modeling method, device and readable storage medium.

背景技术Background technique

随着金融科技，尤其是互联网科技金融的不断发展，越来越多的技术(如分布式、区块链Blockchain、人工智能等)应用在金融领域，但金融业也对技术提出了更高的要求，如对金融业对应待办事项的分发也有更高的要求。With the continuous development of financial technology, especially Internet technology finance, more and more technologies (such as distributed, blockchain, artificial intelligence, etc.) are applied in the financial field, but the financial industry also puts forward higher requirements for technology. Requirements, such as the distribution of corresponding to-do items in the financial industry, also have higher requirements.

随着计算机软件和人工智能的不断发展，联邦学习建模发展的也越来越成熟，目前，联邦学习建模的各参与者通常将己方的加密模型参数反馈至联邦学习建模的协调者，进而协调方者各加密模型参数进行聚合，并将聚合后聚合参数反馈至各参与者，以进行联邦学习建模，但是，若有若恶意参与方在训练过程中提供虚假的加密模型参数，会直接影响联邦学习建模获得的联邦模型的整体模型质量，甚至会导致整个联邦学习建模过程失效，进而使得联邦学习建模的效率和精确度偏低。With the continuous development of computer software and artificial intelligence, the development of federated learning modeling has become more and more mature. At present, each participant of federated learning modeling usually feeds back its own encrypted model parameters to the coordinator of federated learning modeling. Then the coordinator aggregates the parameters of each encryption model, and feeds back the aggregated parameters to the participants for federated learning modeling. However, if a malicious participant provides false encryption model parameters during the training process, it will It directly affects the overall model quality of the federated model obtained by federated learning modeling, and even leads to the failure of the entire federated learning modeling process, which in turn makes the efficiency and accuracy of federated learning modeling low.

发明内容SUMMARY OF THE INVENTION

本申请的主要目的在于提供一种联邦学习建模方法、设备及可读存储介质，旨在解决现有技术中联邦学习建模效率低且精确度差的技术问题。The main purpose of this application is to provide a federated learning modeling method, device and readable storage medium, which aims to solve the technical problems of low efficiency and poor accuracy of federated learning modeling in the prior art.

为实现上述目的，本申请提供一种联邦学习建模方法，所述联邦学习建模方法应用于第一设备，所述联邦学习建模方法包括：To achieve the above purpose, the present application provides a federated learning modeling method, the federated learning modeling method is applied to the first device, and the federated learning modeling method includes:

接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数；receiving encryption model parameters sent by each second device and verification parameters corresponding to the encryption model parameters;

基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果；Based on each of the verification parameters, zero-knowledge verification is performed on each of the encryption model parameters, so as to determine false encryption model parameters in each of the encryption model parameters, and a zero-knowledge verification result is obtained;

基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模。Based on the zero-knowledge verification result and each of the encryption model parameters, each of the second devices is coordinated to perform federated learning modeling.

可选地，所述基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模的步骤包括：Optionally, the step of coordinating each of the second devices to perform federated learning modeling based on the zero-knowledge verification result and each of the encryption model parameters includes:

基于所述零知识验证结果，在各所述加密模型参数中剔除所述虚假加密模型参数，获得各可信模型参数；Based on the zero-knowledge verification result, the false encryption model parameters are removed from the encryption model parameters to obtain each trusted model parameter;

对各所述可信模型参数进行聚合处理，获得聚合参数，并将所述聚合参数分别反馈至各所述第二设备，以供各所述第二设备更新各自的本地训练模型，直至所述本地训练模型达到预设训练结束条件。Perform aggregation processing on each of the trusted model parameters to obtain aggregation parameters, and respectively feed back the aggregation parameters to each of the second devices, so that each of the second devices can update their respective local training models until the The locally trained model reaches the preset training end condition.

可选地，所述基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果的步骤包括：Optionally, performing zero-knowledge verification on each of the encryption model parameters based on each of the verification parameters, so as to determine false encryption model parameters in each of the encryption model parameters, the step of obtaining a zero-knowledge verification result includes:

分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果；respectively calculating the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters;

基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数，获得零知识验证结果。Based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results, it is respectively verified whether each of the encryption model parameters is a false encryption model parameter, and a zero-knowledge verification result is obtained.

可选地，所述验证参数包括验证模型参数和验证随机参数，Optionally, the verification parameters include verification model parameters and verification random parameters,

所述分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果的步骤包括：The step of separately calculating the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters includes:

基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果；Based on preset verification challenge parameters, the validity of each of the encryption model parameters is verified to obtain each of the first zero-knowledge verification results;

基于预设协调方公钥和各所述验证随机参数，对各所述验证模型参数进行加密处理，获得各所述第二零知识验证结果。Based on the preset coordinator public key and each of the verification random parameters, encryption processing is performed on each of the verification model parameters to obtain each of the second zero-knowledge verification results.

可选地，所述预设验证挑战参数包括第一验证挑战参数和第二验证挑战参数；Optionally, the preset verification challenge parameters include a first verification challenge parameter and a second verification challenge parameter;

所述基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果的步骤包括：The step of performing legality verification on each of the encryption model parameters based on the preset verification challenge parameters, and obtaining each of the first zero-knowledge verification results includes:

分别将所述第一验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第一幂操作结果；Perform an exponentiation operation on the first verification challenge parameter and each of the encryption model parameters, respectively, to obtain a first exponentiation operation result corresponding to each of the encryption model parameters;

分别将所述第二验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第二幂操作结果；Performing an exponentiation operation on the second verification challenge parameter and each of the encryption model parameters, respectively, to obtain a second exponentiation operation result corresponding to each of the encryption model parameters;

基于各所述第一幂操作结果和各所述第二幂操作结果，生成各所述第一零知识验证结果。Each of the first zero-knowledge verification results is generated based on each of the first power operation results and each of the second power operation results.

可选地，所述基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数的步骤包括：Optionally, the step of respectively verifying whether each of the encryption model parameters is a fake encryption model parameter based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results includes:

将各所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果分别进行对比；respectively comparing the first zero-knowledge proof result corresponding to each of the encryption model parameters with the second zero-knowledge proof result;

若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果不一致，则判定所述加密模型参数为所述虚假加密模型参数；If the first zero-knowledge proof result corresponding to the encryption model parameter is inconsistent with the second zero-knowledge proof result, it is determined that the encryption model parameter is the false encryption model parameter;

若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果一致，则判定所述加密模型参数不为所述虚假加密模型参数。If the first zero-knowledge proof result corresponding to the encryption model parameter is consistent with the second zero-knowledge proof result, it is determined that the encryption model parameter is not the false encryption model parameter.

为实现上述目的，本申请还提供一种联邦学习建模方法，所述联邦学习建模方法应用于第二设备，所述联邦学习建模方法包括：To achieve the above purpose, the present application also provides a federated learning modeling method, the federated learning modeling method is applied to the second device, and the federated learning modeling method includes:

获取模型训练参数和第一验证随机参数，并基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行加密处理，获得加密模型参数；Obtaining model training parameters and first verification random parameters, and encrypting the model training parameters based on the first verification random parameters and a preset public key to obtain encrypted model parameters;

基于所述第一验证随机参数、所述模型训练参数和预设验证挑战参数，生成验证模型参数和第二验证随机参数；generating a verification model parameter and a second verification random parameter based on the first verification random parameter, the model training parameter and the preset verification challenge parameter;

将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果；sending the encryption model parameter, the verification model parameter and the second verification random parameter to the first device, so that the first device can perform zero-knowledge verification to obtain a zero-knowledge verification result;

接收所述第一设备基于所述零知识验证结果和所述加密模型参数反馈的聚合参数，并基于所述聚合参数，对所述模型训练参数对应的本地训练模型进行更新，直至所述本地训练模型达到预设训练结束条件。Receive the aggregation parameter fed back by the first device based on the zero-knowledge verification result and the encryption model parameter, and update the local training model corresponding to the model training parameter based on the aggregation parameter until the local training The model reaches the preset training end condition.

可选地，所述模型训练参数包括当前模型参数和辅助模型参数，Optionally, the model training parameters include current model parameters and auxiliary model parameters,

所述获取模型训练参数的步骤包括：The step of obtaining model training parameters includes:

对所述模型训练参数对应的本地训练模型进行迭代训练，直至所述本地训练模型达到预设迭代次数阀值，获取所述本地训练模型的所述当前模型参数；Perform iterative training on the local training model corresponding to the model training parameters until the local training model reaches a preset number of iterations threshold, and obtain the current model parameters of the local training model;

获取所述本地训练模型的在前模型参数，并基于所述在前模型参数，生成所述辅助模型参数。Obtaining previous model parameters of the local training model, and generating the auxiliary model parameters based on the previous model parameters.

本申请还提供一种联邦学习建模装置，所述联邦学习建模装置为虚拟装置，且所述联邦学习建模装置应用于第一设备，所述联邦学习建模装置包括：The present application also provides a federated learning modeling device, the federated learning modeling device is a virtual device, and the federated learning modeling device is applied to the first device, and the federated learning modeling device includes:

接收模块，用于接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数；a receiving module, configured to receive the encryption model parameters sent by each second device and the verification parameters corresponding to the encryption model parameters;

零知识验证模块，用于基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果；A zero-knowledge verification module, configured to perform zero-knowledge verification on each of the encryption model parameters based on each of the verification parameters, so as to determine false encryption model parameters in each of the encryption model parameters, and obtain a zero-knowledge verification result;

协调模块，用于基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模。A coordination module, configured to coordinate each of the second devices to perform federated learning modeling based on the zero-knowledge verification result and each of the encryption model parameters.

可选地，所述协调模块包括：Optionally, the coordination module includes:

剔除子模块，用于基于所述零知识验证结果，在各所述加密模型参数中剔除所述虚假加密模型参数，获得各可信模型参数；Elimination sub-module, for eliminating the false encryption model parameters from the encryption model parameters based on the zero-knowledge verification result, to obtain each trusted model parameter;

聚合子模块，用于对各所述可信模型参数进行聚合处理，获得聚合参数，并将所述聚合参数分别反馈至各所述第二设备，以供各所述第二设备更新各自的本地训练模型，直至所述本地训练模型达到预设训练结束条件。an aggregation sub-module, configured to perform aggregation processing on each of the trusted model parameters, obtain aggregation parameters, and feed back the aggregation parameters to each of the second devices, so that each of the second devices can update their local Train the model until the local training model reaches the preset training end condition.

可选地，所述零知识验证模块包括：Optionally, the zero-knowledge verification module includes:

计算子模块，用于分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果；a calculation submodule for calculating the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters;

零知识验证子模块，用于基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数，获得零知识验证结果。The zero-knowledge verification sub-module is configured to verify whether each of the encryption model parameters is a fake encryption model parameter based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results, and obtain a zero-knowledge verification result.

可选地，所述计算子模块包括：Optionally, the calculation submodule includes:

合法性验证单元，用于基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果；a legality verification unit, configured to perform legality verification on each of the encryption model parameters based on preset verification challenge parameters, and obtain each of the first zero-knowledge verification results;

加密单元，用于基于预设协调方公钥和各所述验证随机参数，对各所述验证模型参数进行加密处理，获得各所述第二零知识验证结果。An encryption unit, configured to perform encryption processing on each of the verification model parameters based on the preset coordinator public key and each of the verification random parameters to obtain each of the second zero-knowledge verification results.

可选地，所述合法性验证单元包括：Optionally, the legality verification unit includes:

第一幂操作子单元，用于分别将所述第一验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第一幂操作结果；a first exponentiation operation subunit, configured to perform an exponentiation operation on the first verification challenge parameter and each of the encryption model parameters, respectively, to obtain a first exponentiation operation result corresponding to each of the encryption model parameters;

第二幂操作子单元，用于分别将所述第二验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第二幂操作结果；The second exponentiation operation subunit is used to perform an exponentiation operation on the second verification challenge parameter and each of the encryption model parameters, respectively, to obtain a second exponentiation operation result corresponding to each of the encryption model parameters;

生成子单元，用于基于各所述第一幂操作结果和各所述第二幂操作结果，生成各所述第一零知识验证结果。A generating subunit is configured to generate each of the first zero-knowledge verification results based on each of the first power operation results and each of the second power operation results.

可选地，所述零知识验证子模块包括：Optionally, the zero-knowledge verification submodule includes:

对比单元，用于将各所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果分别进行对比；a comparison unit, configured to compare the first zero-knowledge proof result corresponding to each of the encryption model parameters with the second zero-knowledge proof result;

第一判定单元，用于若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果不一致，则判定所述加密模型参数为所述虚假加密模型参数；a first determination unit, configured to determine that the encryption model parameter is the false encryption model parameter if the first zero-knowledge proof result corresponding to the encryption model parameter is inconsistent with the second zero-knowledge proof result;

第二判定单元，用于若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果一致，则判定所述加密模型参数不为所述虚假加密模型参数。A second determination unit, configured to determine that the encryption model parameter is not the fake encryption model parameter if the first zero-knowledge proof result corresponding to the encryption model parameter is consistent with the second zero-knowledge proof result.

为实现上述目的，本申请还提供一种联邦学习建模装置，所述联邦学习建模装置为虚拟装置，且所述联邦学习建模装置应用于第二设备，所述联邦学习建模装置包括：In order to achieve the above purpose, the present application also provides a federated learning modeling device, the federated learning modeling device is a virtual device, and the federated learning modeling device is applied to the second device, and the federated learning modeling device includes: :

加密模块，用于获取模型训练参数和第一验证随机参数，并基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行加密处理，获得加密模型参数；an encryption module, configured to obtain model training parameters and first verification random parameters, and based on the first verification random parameters and a preset public key, perform encryption processing on the model training parameters to obtain encrypted model parameters;

生成模块，用于基于所述第一验证随机参数、所述模型训练参数和预设验证挑战参数，生成验证模型参数和第二验证随机参数；a generation module, configured to generate a verification model parameter and a second verification random parameter based on the first verification random parameter, the model training parameter and the preset verification challenge parameter;

发送模块，用于将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果；a sending module, configured to send the encryption model parameter, the verification model parameter and the second verification random parameter to the first device, so that the first device can perform zero-knowledge verification to obtain a zero-knowledge verification result;

模型更新模块，用于接收所述第一设备基于所述零知识验证结果和所述加密模型参数反馈的聚合参数，并基于所述聚合参数，对所述模型训练参数对应的本地训练模型进行更新，直至所述本地训练模型达到预设训练结束条件。A model update module, configured to receive the aggregation parameters fed back by the first device based on the zero-knowledge verification result and the encryption model parameters, and update the local training model corresponding to the model training parameters based on the aggregation parameters , until the local training model reaches the preset training end condition.

可选地，所述加密模块包括：Optionally, the encryption module includes:

获取子模块，用于对所述模型训练参数对应的本地训练模型进行迭代训练，直至所述本地训练模型达到预设迭代次数阀值，获取所述本地训练模型的所述当前模型参数；an acquisition submodule, configured to perform iterative training on the local training model corresponding to the model training parameters, until the local training model reaches a preset number of iterations threshold, and acquire the current model parameters of the local training model;

生成子模块，用于获取所述本地训练模型的在前模型参数，并基于所述在前模型参数，生成所述辅助模型参数。A generating submodule is configured to acquire the previous model parameters of the local training model, and generate the auxiliary model parameters based on the previous model parameters.

本申请还提供一种联邦学习建模设备，所述联邦学习建模设备为实体设备，所述联邦学习建模设备包括：存储器、处理器以及存储在所述存储器上并可在所述处理器上运行的所述联邦学习建模方法的程序，所述联邦学习建模方法的程序被处理器执行时可实现如上述的联邦学习建模方法的步骤。The present application also provides a federated learning modeling device, the federated learning modeling device is an entity device, and the federated learning modeling device includes: a memory, a processor, and a device stored on the memory and available on the processor The program of the federated learning modeling method that runs on the federated learning modeling method, when the program of the federated learning modeling method is executed by the processor, can implement the steps of the federated learning modeling method as described above.

本申请还提供一种可读存储介质，所述可读存储介质上存储有实现联邦学习建模方法的程序，所述联邦学习建模方法的程序被处理器执行时实现如上述的联邦学习建模方法的步骤。The present application also provides a readable storage medium on which a program for implementing the federated learning modeling method is stored, and when the program for the federated learning modeling method is executed by a processor, the above-mentioned federated learning modeling method is implemented. steps of the modulo method.

本申请通过接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数，进而基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果，进而基于所述零知识验证结果，协调各所述第二设备进行联邦学习建模。也即，本申请在接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数之后，基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果，进一步地，基于所述零知识验证结果，可在各所述加密模型参数中剔除所述虚假加密模型参数，以协调各所述第二设备进行联邦学习建模。也即，本申请提供了一种基于零知识证明在各所述本地模型中确定虚假加密模型参数的方法，进而当有若恶意参与方在训练过程中提供虚假加密模型参数时，可准确识别并剔除虚假加密模型参数，进而避免了基于混合有虚假加密模型参数的各加密模型参数进行联邦学习建模的情况发生，进而提高了通过联邦学习建模获得的联邦模型的整体模型质量，进而提高了联邦学习建模的效率和精确度，进而解决了联邦学习建模效率低且精确度差的技术问题。The present application receives the encryption model parameters sent by each second device and the verification parameters corresponding to the encryption model parameters, and further performs zero-knowledge verification on the encryption model parameters based on the verification parameters, so as to perform zero-knowledge verification on each of the encryption model parameters. A false encryption model parameter is determined in the encryption model parameters, a zero-knowledge verification result is obtained, and based on the zero-knowledge verification result, each of the second devices is coordinated to perform federated learning modeling. That is, after receiving the encryption model parameters sent by each second device and the verification parameters corresponding to the encryption model parameters, the present application performs zero-knowledge verification on each of the encryption model parameters based on the verification parameters, so that the False encryption model parameters are determined in each of the encryption model parameters, and a zero-knowledge verification result is obtained. The second device performs federated learning modeling. That is, the present application provides a method for determining false encryption model parameters in each of the local models based on zero-knowledge proof, so that when a malicious participant provides false encryption model parameters in the training process, it can accurately identify and Eliminate the false encryption model parameters, thus avoiding the occurrence of federated learning modeling based on the encryption model parameters mixed with false encryption model parameters, thereby improving the overall model quality of the federated model obtained through federated learning modeling, thereby improving the The efficiency and accuracy of federated learning modeling, and then solve the technical problems of low efficiency and poor accuracy of federated learning modeling.

附图说明Description of drawings

此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本申请的实施例，并与说明书一起用于解释本申请的原理。The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description serve to explain the principles of the application.

为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，对于本领域普通技术人员而言，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. In other words, on the premise of no creative labor, other drawings can also be obtained from these drawings.

图1为本申请联邦学习建模方法第一实施例的流程示意图；FIG. 1 is a schematic flowchart of the first embodiment of the federated learning modeling method of the present application;

图2为本申请联邦学习建模方法第二实施例的流程示意图；FIG. 2 is a schematic flowchart of the second embodiment of the federated learning modeling method of the present application;

图3为本申请联邦学习建模方法第三实施例的流程示意图；FIG. 3 is a schematic flowchart of the third embodiment of the federated learning modeling method of the present application;

图4为本申请实施例方案涉及的硬件运行环境的设备结构示意图。FIG. 4 is a schematic diagram of a device structure of a hardware operating environment involved in the solution of the embodiment of the present application.

本申请目的实现、功能特点及优点将结合实施例，参照附图做进一步说明。The realization, functional features and advantages of the present application will be further described with reference to the accompanying drawings in conjunction with the embodiments.

具体实施方式Detailed ways

应当理解，此处所描述的具体实施例仅用以解释本申请，并不用于限定本申请。It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.

本申请实施例提供一种联邦学习建模方法，在本申请联邦学习建模方法的第一实施例中，所述联邦学习建模方法应用于第一设备，参照图1，所述联邦学习建模方法包括：An embodiment of the present application provides a federated learning modeling method. In the first embodiment of the federated learning modeling method of the present application, the federated learning modeling method is applied to a first device. Referring to FIG. 1 , the federated learning modeling method is applied to the first device. Modular methods include:

步骤S10，接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数；Step S10, receiving encryption model parameters sent by each second device and verification parameters corresponding to the encryption model parameters;

在本实施例中，需要说明的是，在进行联邦学习建模之前，所述第一设备与各所述第二设备进行协商交互，确定标准验证挑战参数，其中，所述标准验证挑战参数的数量可在协商交互时确定。所述第一设备为进行联邦学习建模的协调方，用于协调各所述第二设备进行联邦学习建模，所述第二设备为进行联邦学习建模的参与方，所述加密模型参数为基于同态加密算法进行加密的模型训练参数，例如，假设第二设备持有的参与方公钥为P，用于进行同态加密的第一验证随机参数为r₁，基于同态加密算法对模型训练参数m进行加密后的加密模型参数h_m＝Enc(P，m，r₁)，其中，Enc为同态加密符号，进一步地，所述模型训练参数为所述第二设备的本地训练模型的模型参数，例如，假设所述本地训练模型为线性模型，表达式为Y＝β₀+β₁x₁+β₂x₂+…+β_nx_n，则所述模型参数为向量(β₀+β₁+β₂+…+β_n)。In this embodiment, it should be noted that, before the federated learning modeling is performed, the first device negotiates and interacts with each of the second devices to determine standard verification challenge parameters, wherein the standard verification challenge parameters are The amount can be determined when the interaction is negotiated. The first device is a coordinator for federated learning modeling, and is used to coordinate each of the second devices to perform federated learning modeling, the second device is a participant in federated learning modeling, and the encrypted model parameters is a model training parameter for encryption based on a homomorphic encryption algorithm, for example, assuming that the public key of the participant held by the second device is P, the first verification random parameter used for homomorphic encryption is r₁ , based on the homomorphic encryption algorithm The encrypted model parameter h_m =Enc(P, m, r₁ ) after encrypting the model training parameter m, where Enc is a homomorphic encryption symbol, and further, the model training parameter is the local The model parameters of the training model. For example, if the local training model is a linear model and the expression is Y=β₀ +β₁ x₁ +β₂ x₂ +...+β_n x_n , the model parameters are vectors (β₀ +β₁ +β₂ +…+β_n ).

另外地，需要说明的是，所述验证参数为用于进行零知识证明的参数，其中，所述验证参数包括第二验证随机参数和验证模型参数，其中，所述验证模型参数为所述第二设备基于模型训练参数和验证挑战参数生成的，例如，假设所述验证挑战参数包括参数x₁和x₂，所述模型训练参数为m，则所述验证模型参数为n＝m*x₁+m*x₂，进一步地，所述第二验证随机参数为所述第二设备基于所述第一验证随机参数和所述验证挑战参数生成的，例如，假设所述第一验证随机参数为r₁，所述验证挑战参数包括参数x₁和参数x₂，则所述第二验证随机参数r₂＝r₁^x1*r₁^x2。In addition, it should be noted that the verification parameter is a parameter for performing zero-knowledge proof, wherein the verification parameter includes a second verification random parameter and a verification model parameter, wherein the verification model parameter is the first verification model parameter. The second device is generated based on model training parameters and verification challenge parameters. For example, if the verification challenge parameters include parameters x₁ and x₂ , the model training parameter is m, then the verification model parameter is n=m*x₁ +m*x₂ , further, the second verification random parameter is generated by the second device based on the first verification random parameter and the verification challenge parameter. For example, it is assumed that the first verification random parameter is r₁ , the verification challenge parameter includes a parameter x₁ and a parameter x₂ , then the second verification random parameter r₂ =r₁^x1 *r₁^x2 .

另外地，需要说明的是，各所述参与方中的恶意参与方可通过更改对模型训练参数的同态加密的加密参数，以达到生成虚假加密模型参数的目的，而由于协调方在进行联邦建模时，通常直接对各所述第二设备发送的加密模型参数进行聚合处理，获得聚合参数，若各所述加密模型参数中存在虚假加密模型参数，将影响联邦模型训练的效率和精确度，例如，假设第二设备A发送的加密模型参数为5a，当所述第二设备B为恶意参与方，则发送的虚假加密模型参数为100b，所述聚合处理为加权平均，则聚合处理后获得的第一聚合参数为(5a+100b)/2，当所述第二设备B不为恶意参与方时，则发送的加密模型参数为5b，则聚合处理后获得的第二聚合参数为(5a+5b)/2，由此可见，若各第二设备中存在恶意参与方，第一设备在对各第二设备发送的加密模型参数进行聚合处理获得的第二聚合参数和在无恶意参与方时获得的第一集合参数的差距极大，进而若各所述第二设备中存在恶意参与方，将极大程度上影响联邦模型训练的效率和精确度。In addition, it should be noted that a malicious participant among the participating parties can change the encryption parameters of the homomorphic encryption of the model training parameters to achieve the purpose of generating false encryption model parameters. During modeling, the encryption model parameters sent by each of the second devices are usually directly aggregated to obtain aggregation parameters. If there are false encryption model parameters in each of the encryption model parameters, the efficiency and accuracy of federated model training will be affected. , for example, assuming that the encryption model parameter sent by the second device A is 5a, when the second device B is a malicious participant, the false encryption model parameter sent is 100b, and the aggregation process is a weighted average, then after the aggregation process The obtained first aggregation parameter is (5a+100b)/2. When the second device B is not a malicious participant, the sent encryption model parameter is 5b, and the second aggregation parameter obtained after the aggregation process is ( 5a+5b)/2, it can be seen that if there is a malicious participant in each second device, the first device aggregates the encryption model parameters sent by each second device. The difference between the first set of parameters obtained by the party is very large, and if there is a malicious participant in each of the second devices, the efficiency and accuracy of the federated model training will be greatly affected.

另外地，需要说明的是，本实施例中所述第一设备和第二设备均基于同态加密算法进行加密，其中，在一种可实施的方案中，所述同态加密算法应满足以下性质：In addition, it should be noted that in this embodiment, the first device and the second device are encrypted based on a homomorphic encryption algorithm, wherein, in an implementable solution, the homomorphic encryption algorithm should satisfy the following nature:

C＝Enc(PK，m，r)，且对于C₁＝Enc(PK，m₁，r₁))和C₂＝Enc(PK，m₂，r₂)，满足：

其中，C、C₁和C₂均为加密之后的待加密参数，PK为加密的密钥，m、m₁和m₂为待加密参数，r、r₁和r₂为加密所需的随机数。C=Enc(PK, m, r), and for C₁ =Enc(PK, m₁ , r₁ )) and C₂ =Enc(PK, m₂ , r₂ ), satisfy:

Among them, C, C₁ and C₂ are the parameters to be encrypted after encryption, PK is the encrypted key, m, m₁ and m₂ are the parameters to be encrypted, and r, r₁ and r₂ are random data required for encryption. number.

步骤S20，基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果；Step S20, based on each of the verification parameters, perform zero-knowledge verification on each of the encryption model parameters, so as to determine false encryption model parameters in each of the encryption model parameters, and obtain a zero-knowledge verification result;

在本实施例中，基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果，具体地，基于各所述验证参数，分别计算各所述加密模型参数对应的第一零知识证明结果和第二零知识证明结果，进而基于各所述加密模型参数对应的第一零知识证明结果和第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果，其中，所述零知识验证结果为各所述加密模型参数是否为虚假加密模型参数的验证结果，所述虚假加密模型参数为恶意参与方对模型训练参数进行了恶意加密的模型训练参数，例如，通过更改进行同态加密时的加密参数，以实现恶意加密。In this embodiment, based on each of the verification parameters, zero-knowledge verification is performed on each of the encryption model parameters, so as to determine false encryption model parameters in each of the encryption model parameters, and a zero-knowledge verification result is obtained. Specifically, Based on each of the verification parameters, the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the encryption model parameters are calculated respectively, and then based on the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the encryption model parameters The zero-knowledge proof result is to verify whether each of the encryption model parameters is a false encryption model parameter, so as to determine the false encryption model parameters in each of the encryption model parameters, and obtain a zero-knowledge verification result, wherein the zero-knowledge verification result is Whether each of the encryption model parameters is a verification result of a fake encryption model parameter, the fake encryption model parameter is a model training parameter maliciously encrypted by a malicious participant on the model training parameter, for example, encryption when performing homomorphic encryption by changing parameters to implement malicious encryption.

其中，所述基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果的步骤包括：Wherein, based on each of the verification parameters, zero-knowledge verification is performed on each of the encryption model parameters to determine false encryption model parameters in each of the encryption model parameters, and the steps of obtaining a zero-knowledge verification result include:

步骤S21，分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果；Step S21, respectively calculating the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters;

在本实施例中，需要说明的是，所述验证参数包括验证模型参数和第二验证随机参数。In this embodiment, it should be noted that the verification parameters include verification model parameters and second verification random parameters.

分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果，具体地，对于每一所述验证参数均执行以下步骤：Calculate the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters. Specifically, the following steps are performed for each of the verification parameters:

基于预设验证挑战参数，对所述加密模型参数进行同态加和运算，获得第一零知识证明结果，并基于预设协调方公钥、第二验证随机参数，对所述验证模型参数进行同态加密运算，获得第二零知识证明结果，例如，假设预设验证挑战参数为x₁和x₂，加密模型参数为h_m＝Enc(P₁，m，r₁)，其中，P₁为参与方公钥，r₁为第一验证随机参数，m为模型训练参数，且验证模型参数n＝m*x₁+m*x₂，且

进而第一零知识验证结果为

所述第二零知识验证结果为

其中，P₂为协调方公钥，且若各所述第二设备为进行恶意修改，则所述参与方公钥和协调方公钥应当一致。Based on the preset verification challenge parameters, a homomorphic addition operation is performed on the encryption model parameters to obtain a first zero-knowledge proof result, and based on the preset coordinator public key and the second verification random parameter, the verification model parameters are verified. Homomorphic encryption operation to obtain the second zero-knowledge proof result, for example, assuming that the preset verification challenge parameters are x₁ and x₂ , and the encryption model parameters are h_m =Enc(P₁ , m, r₁ ), where P₁ is the public key of the participant, r₁ is the first verification random parameter, m is the model training parameter, and the verification model parameter n=m*x₁ +m*x₂ , and

Then the first zero-knowledge verification result is

The second zero-knowledge verification result is

Wherein, P₂ is the public key of the coordinator, and if each of the second devices is maliciously modified, the public key of the participant and the public key of the coordinator should be consistent.

步骤S22，基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数，获得零知识验证结果。Step S22, based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results, respectively verify whether each of the encryption model parameters is a false encryption model parameter, and obtain a zero-knowledge verification result.

在本实施例中，基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数，获得零知识验证结果，具体地，对于每一所述验证参数对应的第一零知识证明结果和第二零知识证明结果均执行以下步骤：In this embodiment, based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results, it is respectively verified whether each of the encryption model parameters is a false encryption model parameter, and a zero-knowledge verification result is obtained, specifically , the following steps are performed for the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters:

将所述第一零知识证明结果与所述第二零知识证明结果进行比对，判断所述第一零知识证明结果与所述第二零知识证明结果是否一致，若所述第一零知识证明结果与所述第二零知识证明结果一致，则判定所述第二设备在对模型训练参数进行同态加密时，并未进行恶意修改，也即，所述加密模型参数不是虚假加密模型参数，若所述第一零知识证明结果与所述第二零知识证明结果不一致，则判定所述第二设备在对模型训练参数进行同态加密时，进行了恶意修改，也即，所述加密模型参数是虚假加密模型参数。Compare the first zero-knowledge proof result with the second zero-knowledge proof result, and determine whether the first zero-knowledge proof result is consistent with the second zero-knowledge proof result, if the first zero-knowledge proof result is consistent If the proof result is consistent with the second zero-knowledge proof result, it is determined that the second device did not maliciously modify the model training parameters when performing homomorphic encryption, that is, the encrypted model parameters are not fake encryption model parameters , if the first zero-knowledge proof result is inconsistent with the second zero-knowledge proof result, it is determined that the second device has maliciously modified the model training parameters when performing homomorphic encryption, that is, the encryption The model parameters are fake encrypted model parameters.

其中，所述基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数的步骤包括：Wherein, based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results, the step of respectively verifying whether each of the encryption model parameters is a false encryption model parameter includes:

步骤S221，将各所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果分别进行对比；Step S221, comparing the first zero-knowledge proof result corresponding to each of the encryption model parameters with the second zero-knowledge proof result;

在本实施例中，将各所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果分别进行对比，具体地，将各所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果分别进行对比，以分别计算每一所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果之间的差值。In this embodiment, the first zero-knowledge proof result corresponding to each encryption model parameter is compared with the second zero-knowledge proof result. The first zero-knowledge proof result and the second zero-knowledge proof result are compared respectively to calculate the difference between the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the encryption model parameters. difference value.

步骤S222，若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果不一致，则判定所述加密模型参数为所述虚假加密模型参数；Step S222, if the first zero-knowledge proof result corresponding to the encryption model parameter is inconsistent with the second zero-knowledge proof result, determine that the encryption model parameter is the false encryption model parameter;

在本实施例中，若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果不一致，则判定所述加密模型参数为所述虚假加密模型参数，具体地，若所述差值不为0，则判定所述加密模型参数对应的第二设备对模型训练参数进行了恶意加密，进而判定所述加密模型参数是虚假加密模型参数，进而为所述加密模型参数赋予虚假标识，以标识所述加密模型参数为虚假加密模型参数。In this embodiment, if the first zero-knowledge proof result corresponding to the encryption model parameter is inconsistent with the second zero-knowledge proof result, it is determined that the encryption model parameter is the false encryption model parameter, specifically , if the difference is not 0, then it is determined that the second device corresponding to the encryption model parameters maliciously encrypts the model training parameters, and then it is determined that the encryption model parameters are false encryption model parameters, and then the encryption model The parameter is given a false identification to identify the encryption model parameter as a false encryption model parameter.

步骤S223，若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果一致，则判定所述加密模型参数不为所述虚假加密模型参数。Step S223, if the first zero-knowledge proof result corresponding to the encryption model parameter is consistent with the second zero-knowledge proof result, it is determined that the encryption model parameter is not the fake encryption model parameter.

在本实施例中，若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果一致，则判定所述加密模型参数不为所述虚假加密模型参数，具体地，若所述差值为0，则判定所述加密模型参数对应的第二设备未对模型训练参数进行恶意加密，进而判定所述加密模型参数不是虚假加密模型参数，进而为所述加密模型参数赋予可信标识，以标识所述加密模型参数为可信模型参数。In this embodiment, if the first zero-knowledge proof result corresponding to the encryption model parameter is consistent with the second zero-knowledge proof result, it is determined that the encryption model parameter is not the false encryption model parameter, specifically ground, if the difference is 0, then it is determined that the second device corresponding to the encryption model parameter does not maliciously encrypt the model training parameters, and then it is determined that the encryption model parameter is not a false encryption model parameter, and then the encryption model The parameter is given a trusted identifier to identify the encryption model parameter as a trusted model parameter.

步骤S30，基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模。Step S30, based on the zero-knowledge verification result and each of the encryption model parameters, coordinate each of the second devices to perform federated learning modeling.

在本实施例中，需要说明的是，所述零知识验证结果包括各所述加密模型参数对应的标定标识，其中，所述标定标识为标识所述加密模型参数是否为虚假加密模型参数的标识。In this embodiment, it should be noted that the zero-knowledge verification result includes a calibration identifier corresponding to each of the encryption model parameters, wherein the calibration identifier is an identifier that identifies whether the encryption model parameter is a false encryption model parameter .

基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模，具体地，基于各所述标定标识，在各所述加密模型参数中剔除虚假加密模型参数，获得各可信模型参数，进而对各所述可信模型参数进行聚合处理，获得聚合参数，进而基于各所述聚合模型参数，协调各所述第二设备进行联邦学习建模。Based on the zero-knowledge verification result and each of the encryption model parameters, coordinate each of the second devices to perform federated learning modeling, and specifically, based on each of the calibration identifiers, eliminate false encryption models from each of the encryption model parameters parameters, obtain each trusted model parameter, and then perform aggregation processing on each of the trusted model parameters to obtain the aggregation parameter, and then coordinate each of the second devices to perform federated learning modeling based on each of the aggregation model parameters.

其中，所述基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模的步骤包括：Wherein, the step of coordinating each of the second devices to perform federated learning modeling based on the zero-knowledge verification result and each of the encryption model parameters includes:

步骤S31，基于所述零知识验证结果，在各所述加密模型参数中剔除所述虚假加密模型参数，获得各可信模型参数；Step S31, based on the zero-knowledge verification result, remove the false encryption model parameters from each of the encryption model parameters, and obtain each trusted model parameter;

在本实施例中，需要说明的是，协调方在发现虚假加密模型参数后，可以按照预设激励机制处罚所述虚假加密模型参数对应的恶意参与方或者取消恶意参与方后续参与联邦学习建模的资格。In this embodiment, it should be noted that, after discovering the false encryption model parameters, the coordinator can punish the malicious participant corresponding to the false encryption model parameter according to the preset incentive mechanism or cancel the malicious participant's subsequent participation in federated learning modeling Qualifications.

步骤S32，对各所述可信模型参数进行聚合处理，获得聚合参数，并将所述聚合参数分别反馈至各所述第二设备，以供各所述第二设备更新各自的本地训练模型，直至所述本地训练模型达到预设训练结束条件。Step S32, performing aggregation processing on each of the trusted model parameters to obtain aggregation parameters, and feeding back the aggregation parameters to each of the second devices, so that each of the second devices can update their respective local training models, Until the local training model reaches the preset training end condition.

在本实施例中，对各所述可信模型参数进行聚合处理，获得聚合参数，并将所述聚合参数分别反馈至各所述第二设备，以供各所述第二设备更新各自的本地训练模型，直至所述本地训练模型达到预设训练结束条件，具体地，基于预设聚合处理规则，对各所述可信模型参数进行聚合处理，获得聚合参数，其中，所述预设集合处理规则包括加权求平均、求和等，进一步地，将各所述聚合参数分别发送至各所述第二设备，以供各所述第二设备基于所述参与方公钥对应的参与方私钥，对所述聚合参数进行解密，获得解密聚合参数，并基于解密聚合参数，对己方持有的本地训练模型进行更新，获得更新后的本地训练模型，并判断更新后的本地训练模型是否达到预设训练结束条件，若更新后的本地训练模型达到预设训练结束条件，则判定完成所述联邦学习建模的任务，若更新后的本地训练模型未达到预设训练结束条件，则重新对所述本地训练模型进行迭代训练，直至所述本地模型达到预设迭代次数阀值，则重新获取所述本地训练模型的模型训练参数，并将所述模型训练参数重新加密发送至所述协调方，以重新进行联邦，直至所述本地训练模型达到预设训练结束条件，其中，所述训练结束条件包括达到预设最大迭代次数，所述本地训练模型对应的损失函数收敛等。In this embodiment, aggregation processing is performed on each of the trusted model parameters to obtain aggregation parameters, and the aggregation parameters are fed back to each of the second devices, so that each of the second devices can update their local Train the model until the local training model reaches a preset training end condition, specifically, based on a preset aggregation processing rule, perform aggregation processing on each of the trusted model parameters to obtain aggregation parameters, wherein the preset set processing The rules include weighted average, summation, etc. Further, each of the aggregation parameters is sent to each of the second devices, so that each of the second devices can use the participant's private key corresponding to the participant's public key , decrypt the aggregation parameters to obtain the decrypted aggregation parameters, and based on the decrypted aggregation parameters, update the local training model held by oneself, obtain the updated local training model, and determine whether the updated local training model reaches the predetermined level. Set the training end condition. If the updated local training model reaches the preset training end condition, it is determined that the task of federated learning modeling is completed. If the updated local training model does not meet the preset training end condition, the The local training model is iteratively trained until the local model reaches the preset number of iterations threshold, then the model training parameters of the local training model are re-acquired, and the model training parameters are re-encrypted and sent to the coordinator, Re-federation is performed until the local training model reaches a preset training end condition, wherein the training end condition includes reaching a preset maximum number of iterations, the loss function corresponding to the local training model converges, and the like.

进一步地，所述本地训练模型包括风控模型，其中，所述风控模型为用于评估用户贷款风险的机器学习模型，进而当各所述参与方中存在恶意参与方时，则基于所述恶意参与方发送的虚假加密模型参数和各所述参与方中的正常参与方发送的加密模型参数进行聚合，将获得相比于与准确聚合参数相差极大的误差聚合参数，进而各第二设备基于误差聚合参数更新风控模型，将降低风控模型对用户贷款风险评估的准确性，进而若基于本申请中的联邦学习建模方法，可从各所述参与方发送的加密模型参数中筛选并剔除虚假加密模型参数，获得可信加密模型参数，进而在整个联邦学习建模过程中，所述风控模型总是基于对各可信加密模型参数进行聚合处理获得的聚合参数进行更新的，进而使得风控模型对用户贷款风险的评估更加准确，也即，提高了风控模型的贷款风险评估准确率。Further, the local training model includes a risk control model, wherein the risk control model is a machine learning model for evaluating the user's loan risk, and when there is a malicious participant in each of the participants, based on the The false encryption model parameters sent by the malicious participants are aggregated with the encryption model parameters sent by the normal participants among the participants, and the error aggregation parameters that are greatly different from the accurate aggregation parameters will be obtained. Updating the risk control model based on the error aggregation parameters will reduce the accuracy of the risk control model for user loan risk assessment, and if based on the federated learning modeling method in this application, it can be screened from the encrypted model parameters sent by the participants. And remove the false encryption model parameters to obtain trusted encryption model parameters, and then in the entire federated learning modeling process, the risk control model is always updated based on the aggregation parameters obtained by the aggregation processing of each trusted encryption model parameter, This makes the risk control model more accurate in evaluating the user's loan risk, that is, improving the loan risk evaluation accuracy rate of the risk control model.

本实施例通过接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数，进而基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果，进而基于所述零知识验证结果，协调各所述第二设备进行联邦学习建模。也即，本实施例在接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数之后，基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果，进一步地，基于所述零知识验证结果，可在各所述加密模型参数中剔除所述虚假加密模型参数，以协调各所述第二设备进行联邦学习建模。也即，本实施例提供了一种基于零知识证明在各所述本地模型中确定虚假加密模型参数的方法，进而当有若恶意参与方在训练过程中提供虚假加密模型参数时，可准确识别并剔除虚假加密模型参数，进而避免了基于混合有虚假加密模型参数的各加密模型参数进行联邦学习建模的情况发生，进而提高了通过联邦学习建模获得的联邦模型的整体模型质量，进而提高了联邦学习建模的效率和精确度，进而解决了联邦学习建模效率低且精确度差的技术问题。In this embodiment, the encryption model parameters sent by each second device and the verification parameters corresponding to the encryption model parameters are received, and based on each of the verification parameters, zero-knowledge verification is performed on each of the encryption model parameters, so that the The false encryption model parameters are determined from the encryption model parameters, the zero-knowledge verification result is obtained, and then based on the zero-knowledge verification result, each of the second devices is coordinated to perform federated learning modeling. That is, in this embodiment, after receiving the encryption model parameters sent by each second device and the verification parameters corresponding to the encryption model parameters, based on the verification parameters, zero-knowledge verification is performed on each of the encryption model parameters to obtain Determine fake encryption model parameters in each of the encryption model parameters, and obtain a zero-knowledge verification result. Further, based on the zero-knowledge verification result, the fake encryption model parameters can be eliminated from each of the encryption model parameters to coordinate Each of the second devices performs federated learning modeling. That is, this embodiment provides a method for determining false encryption model parameters in each of the local models based on zero-knowledge proof, so that when a malicious participant provides false encryption model parameters in the training process, it can be accurately identified. And eliminate the false encryption model parameters, thus avoiding the occurrence of federated learning modeling based on the encryption model parameters mixed with false encryption model parameters, thereby improving the overall model quality of the federated model obtained through federated learning modeling, thereby improving The efficiency and accuracy of federated learning modeling are improved, and the technical problems of low efficiency and poor accuracy of federated learning modeling are solved.

进一步地，参照图2，基于本申请中第一实施例，在本申请的另一实施例中，所述验证参数包括验证模型参数和验证随机参数，Further, referring to FIG. 2 , based on the first embodiment of the present application, in another embodiment of the present application, the verification parameters include verification model parameters and verification random parameters,

所述分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果的步骤包括：The step of separately calculating the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters includes:

步骤S211，基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果；Step S211, based on preset verification challenge parameters, perform legality verification on each of the encryption model parameters, and obtain each of the first zero-knowledge verification results;

在本实施例中，需要说明的是，所述预设验证挑战参数包括第一验证挑战参数和第二验证挑战参数，所述加密模型参数包括第一加密模型参数和第二加密模型参数，其中，所述第一加密模型参数为所述第二设备对当前模型参数进行同态加密之后的加密参数，所述第二加密模型参数为所述第二设备对在前模型参数进行同态加密之后的加密参数，其中，所述当前模型参数为在进行本轮联邦时，所述本地训练模型达到预设训练迭代次数阀值时提取的模型参数，所述在前模型参数为基于本轮联邦之前的在前轮次联邦是的模型参数，例如，取前三轮联邦对应的历史模型参数，并对各所述历史模型参数进行加权求平均，获得所述在前模型参数，例如，假设，所述当前模型参数为m，所述在前模型参数为m₀，则第一加密模型参数h₀＝Enc(P，m₀，r₁)，其中，P为参与方公钥，r₁为第一验证随机参数，第二加密模型参数h_m＝Enc(P，m，r₂)，其中，r₂为第二验证随机参数。In this embodiment, it should be noted that the preset verification challenge parameters include a first verification challenge parameter and a second verification challenge parameter, and the encryption model parameters include a first encryption model parameter and a second encryption model parameter, wherein , the first encryption model parameter is the encryption parameter after the second device performs homomorphic encryption on the current model parameter, and the second encryption model parameter is after the second device performs homomorphic encryption on the previous model parameter encryption parameters, wherein the current model parameter is the model parameter extracted when the local training model reaches the preset training iteration threshold when the current round of federation is performed, and the previous model parameter is based on the current round of federation. The model parameters of the previous rounds of federations are, for example, take the historical model parameters corresponding to the previous three rounds of federations, and perform a weighted average of the historical model parameters to obtain the previous model parameters, for example, suppose that all If the current model parameter is m, and the previous model parameter is m₀ , then the first encryption model parameter h₀ =Enc(P, m₀ , r₁ ), where P is the public key of the participant, and r₁ is the first encryption model parameter. A verification random parameter, the second encryption model parameter h_m =Enc(P, m, r₂ ), where r₂ is the second verification random parameter.

基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果，具体地，对于每一所述加密模型参数均执行以下步骤：Based on the preset verification challenge parameters, the validity of each of the encryption model parameters is verified to obtain each of the first zero-knowledge verification results. Specifically, the following steps are performed for each of the encryption model parameters:

基于所述第一验证挑战参数和所述第二验证挑战参数，分别对所述第一加密模型参数和第二加密模型参数进行幂操作并求和，获得第一零知识验证结果，例如，假设第一加密模型参数h₀＝Enc(P，m₀，r₁)，第二加密模型参数h_m＝Enc(P，m，r₂)，第一验证挑战参数为x₁，第二验证挑战参数为x₂，则所述第一零知识证明结果

进而基于同态加密算法的性质，可得

其中，P为参与方公钥，x₁为第一验证挑战参数，x₂为第二验证挑战参数，r₁为第一验证随机参数，r₂为第二验证随机参数，所述当前模型参数为m，所述在前模型参数为m₀。Based on the first verification challenge parameter and the second verification challenge parameter, exponentiate and sum the first encryption model parameter and the second encryption model parameter, respectively, to obtain a first zero-knowledge verification result, for example, assuming The first encryption model parameter h₀ =Enc(P, m₀ , r₁ ), the second encryption model parameter h_m =Enc(P, m, r₂ ), the first verification challenge parameter is x₁ , the second verification challenge The parameter is x₂ , then the first zero-knowledge proof result

Based on the properties of the homomorphic encryption algorithm, we can get

Among them, P is the public key of the participant, x₁ is the first verification challenge parameter, x₂ is the second verification challenge parameter, r₁ is the first verification random parameter, r₂ is the second verification random parameter, and the current model parameter is m, and the preceding model parameter is m₀ .

其中，所述预设验证挑战参数包括第一验证挑战参数和第二验证挑战参数；Wherein, the preset verification challenge parameters include a first verification challenge parameter and a second verification challenge parameter;

所述基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果的步骤包括：The step of performing legality verification on each of the encryption model parameters based on the preset verification challenge parameters, and obtaining each of the first zero-knowledge verification results includes:

步骤A10，分别将所述第一验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第一幂操作结果；Step A10, respectively performing an exponentiation operation on the first verification challenge parameter and each of the encryption model parameters to obtain a first exponentiation operation result corresponding to each of the encryption model parameters;

在本实施例中，分别将所述第一验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第一幂操作结果，具体地，对于每一所述加密模型参数，均执行以下步骤：基于所述第一验证挑战参数，对所述第一加密模型参数进行幂操作，获得第一幂操作结果，例如，假设所述第一验证挑战参数为x，所述第一加密模型参数为h，则所述第一幂操作结果为h^x。In this embodiment, an exponentiation operation is performed on the first verification challenge parameter and each of the encryption model parameters to obtain a first exponentiation operation result corresponding to each of the encryption model parameters. model parameters, perform the following steps: based on the first verification challenge parameter, perform an exponentiation operation on the first encryption model parameter to obtain a first exponentiation operation result. For example, assuming that the first verification challenge parameter is x, so If the first encryption model parameter is h, the result of the first power operation is h^x .

步骤A20，分别将所述第二验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第二幂操作结果；Step A20, performing an exponentiation operation on the second verification challenge parameter and each of the encryption model parameters, respectively, to obtain a second exponentiation operation result corresponding to each of the encryption model parameters;

在本实施例中，分别将所述第二验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第二幂操作结果，具体地，对于每一所述加密模型参数，均执行以下步骤：基于所述第二验证挑战参数，对所述第二加密模型参数进行幂操作，获得第二幂操作结果。In this embodiment, the second verification challenge parameter and each of the encryption model parameters are respectively subjected to an exponentiation operation to obtain a second exponentiation operation result corresponding to each of the encryption model parameters. model parameters, the following steps are performed: based on the second verification challenge parameter, perform an exponentiation operation on the second encryption model parameter to obtain a second exponentiation operation result.

步骤A30，基于各所述第一幂操作结果和各所述第二幂操作结果，生成各所述第一零知识验证结果。Step A30: Generate each of the first zero-knowledge verification results based on each of the first power operation results and each of the second power operation results.

在本实施例中，基于各所述第一幂操作结果和各所述第二幂操作结果，生成各所述第一零知识验证结果，具体地，求取所述第一幂操作结果和所述第二幂操作结果的乘积，并将所述乘积作为所述第一零知识验证结果。In this embodiment, each of the first zero-knowledge verification results is generated based on each of the first exponentiation operation results and each of the second exponentiation operation results, and specifically, the first exponentiation operation result and the The product of the second power operation results is obtained, and the product is used as the first zero-knowledge verification result.

步骤S212，基于预设协调方公钥和各所述验证随机参数，对各所述验证模型参数进行加密处理，获得各所述第二零知识验证结果。Step S212: Encrypt each of the verification model parameters based on the preset coordinator public key and each of the verification random parameters to obtain each of the second zero-knowledge verification results.

在本实施例中，需要说明的是，各所述参与方均为合法参与方，进而所述参与方公钥和所述预设协调方公钥一致，所述验证随机参数包括第三验证随机参数，且所述第三验证随机参数为基于第一验证随机参数、第二验证随机参数、第一验证挑战参数和第二验证挑战参数计算获得的，例如，假设第一验证随机参数为r1，第二验证随机参数为r2，第一验证挑战参数为x1，第二验证挑战参数为x2，则第三验证挑战参数

In this embodiment, it should be noted that each of the participants is a legal participant, and the public key of the participant is consistent with the public key of the preset coordinator, and the verification random parameter includes a third verification random parameter, and the third verification random parameter is calculated based on the first verification random parameter, the second verification random parameter, the first verification challenge parameter and the second verification challenge parameter, for example, assuming that the first verification random parameter is r1, The second verification random parameter is r2, the first verification challenge parameter is x1, the second verification challenge parameter is x2, then the third verification challenge parameter

另外地，所述验证模型参数为基于第一验证挑战参数、第二验证挑战参数、当前模型参数和在前模型参数计算获得的，例如，假设所述第一验证挑战参数为x1，第二验证挑战参数为x2，当前模型参数为m，在前模型参数为m₀，则所述验证模型参数为n＝m₀x₁+mx₂。In addition, the verification model parameter is calculated based on the first verification challenge parameter, the second verification challenge parameter, the current model parameter and the previous model parameter. For example, assuming that the first verification challenge parameter is x1, the second verification challenge The challenge parameter is x2, the current model parameter is m, and the previous model parameter is m₀ , then the verification model parameter is n=m₀ x₁ +mx₂ .

基于预设协调方公钥和各所述验证随机参数，对各所述验证模型参数进行加密处理，获得各所述第二零知识验证结果，具体地，对于每一所述验证模型参数均执行以下步骤：Based on the preset coordinator public key and each of the verification random parameters, encryption processing is performed on each of the verification model parameters to obtain each of the second zero-knowledge verification results. Specifically, for each of the verification model parameters, the The following steps:

基于预设协调方公钥和所述第三验证挑战参数，对所述验证模型参数进行同态加密，获得所述第二零知识验证结果，例如，假设第三验证挑战参数

其中，第一验证随机参数为r1，第二验证随机参数为r2，第一验证挑战参数为x1，第二验证挑战参数为x2，所述验证模型参数为n＝m₀x₁+mx₂，预设协调方公钥为P，则

Based on the preset coordinator public key and the third verification challenge parameter, perform homomorphic encryption on the verification model parameters to obtain the second zero-knowledge verification result, for example, assuming the third verification challenge parameter

The first verification random parameter is r1, the second verification random parameter is r2, the first verification challenge parameter is x1, the second verification challenge parameter is x2, and the verification model parameter is n=m₀ x₁ +mx₂ , The default public key of the coordinator is P, then

进一步地，若参与方没有对加密模型参数进行恶意加密，例如，恶意篡改加密算法、恶意篡改加密的参数等，则所述第一零知识证明结果与第二零知识证明结果相同，也即，所述参与方向协调方提供的加密模型参数为可信模型参数，若参与方对加密模型参数进行了恶意加密，则所述第一零知识证明结果与第二零知识证明结果不相同，也即，所述参与方向协调方提供的加密模型参数为虚假加密模型参数。Further, if the participant does not maliciously encrypt the encryption model parameters, for example, maliciously tampering with the encryption algorithm, maliciously tampering with the encrypted parameters, etc., the first zero-knowledge proof result is the same as the second zero-knowledge proof result, that is, The encryption model parameters provided by the participant to the coordinator are trusted model parameters. If the participant maliciously encrypts the encryption model parameters, the first zero-knowledge proof result is different from the second zero-knowledge proof result, that is, , the encryption model parameters provided by the participant to the coordinator are false encryption model parameters.

本实施例通过基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果，进而基于预设协调方公钥和各所述验证随机参数，对各所述验证模型参数进行加密处理，获得各所述第二零知识验证结果。也即，本实施例提供了一种计算第一零知识证明结果和第二零知识证明结果的方法，进而在计算获得第一零知识证明结果和第二零知识证明结果之后，将第一零知识证明结果和第二零知识证明结果仅比对，即可确定加密模型参数是否为虚假加密模型参数，进而为在各所述加密模型参数中确定虚假加密模型参数奠定了基础，进而为解决联邦学习建模效率低且精确度差的技术问题奠定了基础。In this embodiment, based on the preset verification challenge parameters, the validity of each encryption model parameter is verified to obtain each of the first zero-knowledge verification results, and then based on the preset coordinator public key and each of the verification random parameters, Perform encryption processing on each of the verification model parameters to obtain each of the second zero-knowledge verification results. That is, this embodiment provides a method for calculating the first zero-knowledge proof result and the second zero-knowledge proof result, and then after the first zero-knowledge proof result and the second zero-knowledge proof result are obtained by calculation, the first zero-knowledge proof result and the second zero-knowledge proof result are obtained. The knowledge proof result and the second zero-knowledge proof result are only compared to determine whether the encryption model parameters are fake encryption model parameters, which lays a foundation for determining the fake encryption model parameters in each of the encryption model parameters, and further solves the problem of federal Lay the groundwork for the technical problems of learning to model inefficiencies and poor accuracy.

进一步地，参照图3，基于本申请中第一实施例和第二实施例，在本申请的另一实施例中，所述联邦学习建模方法应用于第二设备，所述联邦学习建模方法包括：Further, referring to FIG. 3 , based on the first embodiment and the second embodiment of the present application, in another embodiment of the present application, the federated learning modeling method is applied to a second device, and the federated learning modeling Methods include:

步骤B10，获取模型训练参数和第一验证随机参数，并基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行加密处理，获得加密模型参数；Step B10, obtaining model training parameters and first verification random parameters, and encrypting the model training parameters based on the first verification random parameters and a preset public key, to obtain encrypted model parameters;

在本实施例中，所述联邦学习建模至少包括一轮联邦，且在每一轮联邦中，所述第二设备对本地训练模型进行迭代训练，直至达到预设迭代次数阀值，则将所述本地训练模型的模型参数发送至所述第一设备，并接所述第一设备基于所述模型参数反馈的聚合参数，并基于所述聚合参数，更新所述本地训练模型，并将所述本地训练模型作为下一轮联邦的初始模型，直至所述本地训练模型达到预设训练结束条件，其中，所述预设训练结束条件包括达到最大迭代次数、损失函数收敛等。In this embodiment, the federated learning modeling includes at least one round of federation, and in each round of federation, the second device performs iterative training on the local training model until a preset number of iterations threshold is reached, then The model parameters of the local training model are sent to the first device, and the aggregation parameters fed back by the first device based on the model parameters are received, and based on the aggregation parameters, the local training model is updated, and the The local training model is used as the initial model of the next round of federation until the local training model reaches a preset training end condition, wherein the preset training end condition includes reaching the maximum number of iterations, loss function convergence, and the like.

获取模型训练参数和第一验证随机参数，并基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行加密处理，获得加密模型参数，具体地，当所述本地训练模型达到预设迭代次数阀值时，提取本地训练模型的模型参数作为模型训练参数，并获取第一验证随机参数，进而基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行同态加密处理，获得加密模型参数，例如，假设所述模型训练参数为m，所述第一验证随机参数为r₁，所述预设公钥为P，则所述加密模型参数为h_m＝Enc(P，m，r₁)，其中，Enc为同态加密符号。Obtain model training parameters and first verification random parameters, and encrypt the model training parameters based on the first verification random parameters and the preset public key to obtain encrypted model parameters. Specifically, when the local training model When the preset number of iterations threshold is reached, the model parameters of the local training model are extracted as model training parameters, and the first verification random parameters are obtained, and then based on the first verification random parameters and the preset public key, the model training parameters are Perform homomorphic encryption processing to obtain encryption model parameters, for example, assuming that the model training parameter is m, the first verification random parameter is r₁ , the preset public key is P, and the encryption model parameter is h_m =Enc(P, m, r₁ ), where Enc is a homomorphic encryption symbol.

其中，所述模型训练参数包括当前模型参数和辅助模型参数，Wherein, the model training parameters include current model parameters and auxiliary model parameters,

所述获取模型训练参数的步骤包括：The step of obtaining model training parameters includes:

步骤B11，对所述模型训练参数对应的本地训练模型进行迭代训练，直至所述本地训练模型达到预设迭代次数阀值，获取所述本地训练模型的所述当前模型参数；Step B11, performing iterative training on the local training model corresponding to the model training parameters, until the local training model reaches a preset number of iterations threshold, and acquiring the current model parameters of the local training model;

在本实施例中，需要说明的是，所述当前模型参数为所述本地训练模型在本轮联邦中达到预设迭代次数阀值时的本轮迭代模型参数。In this embodiment, it should be noted that the current model parameters are the iteration model parameters of the current round when the local training model reaches the preset number of iterations threshold in the current round of federation.

步骤B12，获取所述本地训练模型的在前模型参数，并基于所述在前模型参数，生成所述辅助模型参数。Step B12: Obtain the previous model parameters of the local training model, and generate the auxiliary model parameters based on the previous model parameters.

在本实施例中，获取所述本地训练模型的在前模型参数，并基于所述在前模型参数，生成所述辅助模型参数，具体地，获取所述本轮联邦对应的在前轮次联邦的各在前迭代模型参数，并对各所述在前迭代模型参数进行加权求平均，获得所述辅助模型参数，例如，假设各所述在前迭代模型参数为a，b，c，a对应的权重为20％，b对应的权重为30％，c对应的权重为50％，则所述辅助模型参数m₀＝a*20％+b*30％+c*50％。In this embodiment, the previous model parameters of the local training model are acquired, and the auxiliary model parameters are generated based on the previous model parameters. Specifically, the previous round of federations corresponding to the current round of federations are acquired Each preceding iteration model parameter is obtained by weighting and averaging the preceding iteration model parameters to obtain the auxiliary model parameters. For example, assuming that each preceding iteration model parameter is a, b, c, a corresponding to is 20%, the weight corresponding to b is 30%, and the weight corresponding to c is 50%, then the auxiliary model parameter m₀ =a*20%+b*30%+c*50%.

步骤B20，基于所述第一验证随机参数、所述模型训练参数和预设验证挑战参数，生成验证模型参数和第二验证随机参数；Step B20, based on the first verification random parameter, the model training parameter and the preset verification challenge parameter, generate a verification model parameter and a second verification random parameter;

在本实施例中，需要说明的是，在一种可行的实施方案中，所述预设验证挑战参数可以由所述协调方根据各参与方在在前轮次联邦中发送的在各前加密模型参数和预设哈希函数计算得到的，例如，假设存在10个参与方，则对对应的10个所述在前加密模型参数进行自由组合，然后将自由组合的n个结果输入到预设哈希函数中，得到验证挑战参数x₁、x₂。x_n，且具体的预设验证挑战参数x₁、x₂。x_n的生成方式和数量均不做限定。In this embodiment, it should be noted that, in a feasible implementation, the preset verification challenge parameter may be encrypted by the coordinator according to the previous encryption sent by each participant in the previous round of federation Calculated from the model parameters and the preset hash function. For example, if there are 10 participants, the corresponding 10 previously encrypted model parameters are freely combined, and then the n results of the free combination are input to the preset In the hash function, the verification challenge parameters x₁ and x₂ are obtained. x_n , and specific preset verification challenge parameters x₁ , x₂ . The generation method and quantity of x_n are not limited.

基于所述第一验证随机参数、所述模型训练参数和预设验证挑战参数，生成验证模型参数和第二验证随机参数，具体地，对所述第一验证随机参数和所述预设验证挑战参数执行幂操作，获得第二验证随机参数，并基于所述模型训练参数和预设验证挑战参数，生成验证模型参数，例如，假设，假设所述第一验证随机参数为r₁，所述模型训练参数为m，所述预设验证挑战参数为x₁和x₂，所述验证模型参数n＝m*x₁+m*x₂，所述第二验证随机参数Based on the first verification random parameter, the model training parameter and the preset verification challenge parameter, a verification model parameter and a second verification random parameter are generated. Perform a power operation on the parameters to obtain a second verification random parameter, and generate verification model parameters based on the model training parameters and preset verification challenge parameters, for example, assuming that the first verification random parameter is r₁ , the model The training parameter is m, the preset verification challenge parameters are x₁ and x₂ , the verification model parameter n=m*x₁ +m*x₂ , the second verification random parameter

步骤B30，将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果；Step B30, sending the encryption model parameter, the verification model parameter and the second verification random parameter to the first device, so that the first device can perform zero-knowledge verification to obtain a zero-knowledge verification result;

在本实施例中，将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果，具体地，将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至与所述第二设备关联的第一设备，以供所述第一设备基于所述加密模型参数、所述验证模型参数和所述第二验证随机参数，计算第一零知识证明结果和第二零知识证明结果，并基于所述第一零知识证明结果和第二零知识证明结果，确定所述加密模型参数是否为虚假加密模型参数，获得确定结果，并将所述确定结果记录于所述零知识验证结果，其中，所述零知识验证结果包括各所述第二设备对应的确定结果。In this embodiment, the encryption model parameter, the verification model parameter and the second verification random parameter are sent to the first device, so that the first device can perform zero-knowledge verification to obtain a zero-knowledge verification result, Specifically, the encryption model parameters, the verification model parameters and the second verification random parameters are sent to a first device associated with the second device for the first device to use the encryption model parameters based on the encryption model parameters. , the verification model parameters and the second verification random parameters, calculate the first zero-knowledge proof result and the second zero-knowledge proof result, and based on the first zero-knowledge proof result and the second zero-knowledge proof result, determine the determine whether the encryption model parameters are false encryption model parameters, obtain a determination result, and record the determination result in the zero-knowledge verification result, wherein the zero-knowledge verification result includes determination results corresponding to each of the second devices.

步骤B40，接收所述第一设备基于所述零知识验证结果和所述加密模型参数反馈的聚合参数，并基于所述聚合参数，对所述模型训练参数对应的本地训练模型进行更新，直至所述本地训练模型达到预设训练结束条件。Step B40: Receive the aggregation parameters fed back by the first device based on the zero-knowledge verification result and the encryption model parameters, and update the local training model corresponding to the model training parameters based on the aggregation parameters until all the parameters are The local training model reaches the preset training end condition.

在本实施例中，接收所述第一设备基于所述零知识验证结果和所述加密模型参数反馈的聚合参数，并基于所述聚合参数，对所述模型训练参数对应的本地训练模型进行更新，直至所述本地训练模型达到预设训练结束条件，具体地，所述第一设备在获得零知识证明结果后，将基于所述零知识证明结果，在各所述第二设备发送的加密模型参数中剔除虚假加密模型参数，获得各可信模型参数，并对各所述可信模型参进行聚合处理，其中，所述聚合处理包括求和、加权求平均等，获得聚合参数，并将聚合参数分别反馈至各所述第二设备，进而第二设备在结束所述聚合参数后，基于所述预设公钥对应的预设私钥，对所述聚合参数进行解密，获得解密聚合参数，并基于所述解密聚合参数，更新所述本地训练模型，并将更新后的所述本地训练模型作为下一轮联邦的初始模型，直至所述本地训练模型达到预设训练结束条件，其中，所述训练结束条件包括达到最大迭代次数和损失函数收敛等。In this embodiment, the aggregation parameter fed back by the first device based on the zero-knowledge verification result and the encryption model parameter is received, and based on the aggregation parameter, the local training model corresponding to the model training parameter is updated , until the local training model reaches the preset training end condition. Specifically, after the first device obtains the zero-knowledge proof result, based on the zero-knowledge proof result, the encryption model sent by each second device Eliminate false encryption model parameters from the parameters, obtain each trusted model parameter, and perform aggregation processing on each of the trusted model parameters, wherein the aggregation processing includes summation, weighted averaging, etc., to obtain aggregation parameters, and aggregate them. The parameters are respectively fed back to each of the second devices, and after finishing the aggregation parameters, the second device decrypts the aggregation parameters based on the preset private key corresponding to the preset public key to obtain the decrypted aggregation parameters, And based on the decryption aggregation parameters, the local training model is updated, and the updated local training model is used as the initial model of the next round of federation, until the local training model reaches the preset training end condition, wherein the The training end conditions include reaching the maximum number of iterations and convergence of the loss function.

本实施例通过获取模型训练参数和第一验证随机参数，并基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行加密处理，获得加密模型参数，进而基于所述第一验证随机参数、所述模型训练参数和预设验证挑战参数，生成验证模型参数和第二验证随机参数，进而将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果，进而接收所述第一设备基于所述零知识验证结果和所述加密模型参数反馈的聚合参数，并基于所述聚合参数，对所述模型训练参数对应的本地训练模型进行更新，直至所述本地训练模型达到预设训练结束条件。也即，本实施例提供了一种基于零知识证明的联邦学习建模方法，也即，在对模型训练参数加密为加密模型参数时，同时生成验证模型参数和第二验证随参数，进而将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果，进而所述第一设备在各所述第二设备中确定并剔除虚假加密模型参数，进而所述第二设备接收的聚合参数所述第一设备为基于可信的加密模型参数聚合获得的，进而基于所述聚合参数，对本地训练模型进行更新，以完成联邦学习建模，进而避免了基于混合有虚假加密模型参数的各加密模型参数聚合的聚合参数更新本地训练模型，使得本地训练模型难以达到预设训练结束条件且本地训练模型确定低的情况发生，进而提高了联邦学习建模的效率和精确度，进而解决了联邦学习建模效率低且精确度差的技术问题。In this embodiment, the model training parameters and the first verification random parameters are obtained, and based on the first verification random parameters and the preset public key, the model training parameters are encrypted to obtain the encrypted model parameters, and then based on the first verification random parameters and the preset public key, the model training parameters are encrypted. a verification random parameter, the model training parameter and the preset verification challenge parameter, generate a verification model parameter and a second verification random parameter, and then send the encryption model parameter, the verification model parameter and the second verification random parameter to the first device, so that the first device can perform zero-knowledge verification, obtain a zero-knowledge verification result, and then receive the aggregated parameters fed back by the first device based on the zero-knowledge verification result and the encryption model parameters, and based on the For the aggregation parameter, the local training model corresponding to the model training parameter is updated until the local training model reaches a preset training end condition. That is, this embodiment provides a zero-knowledge proof-based federated learning modeling method, that is, when the model training parameters are encrypted into encrypted model parameters, the verification model parameters and the second verification parameters are generated at the same time, and then the The encryption model parameters, the verification model parameters, and the second verification random parameters are sent to the first device for the first device to perform zero-knowledge verification to obtain a zero-knowledge verification result, and then the first device performs zero-knowledge verification. The false encryption model parameters are determined and eliminated in each of the second devices, and then the aggregation parameters received by the second device are obtained by the first device based on the aggregation of trusted encryption model parameters, and then based on the aggregation parameters, The local training model is updated to complete the federated learning modeling, thereby avoiding updating the local training model based on the aggregation parameters of each encryption model parameter aggregation mixed with false encryption model parameters, making it difficult for the local training model to reach the preset training end condition and the local The training model determines that low occurrences occur, thereby improving the efficiency and accuracy of federated learning modeling, thereby solving the technical problems of low efficiency and poor accuracy of federated learning modeling.

参照图4，图4是本申请实施例方案涉及的硬件运行环境的设备结构示意图。Referring to FIG. 4 , FIG. 4 is a schematic diagram of the device structure of the hardware operating environment involved in the solution of the embodiment of the present application.

如图4所示，该联邦学习建模设备可以包括：处理器1001，例如CPU，存储器1005，通信总线1002。其中，通信总线1002用于实现处理器1001和存储器1005之间的连接通信。存储器1005可以是高速RAM存储器，也可以是稳定的存储器(non-volatile memory)，例如磁盘存储器。存储器1005可选的还可以是独立于前述处理器1001的存储设备。As shown in FIG. 4 , the federated learning modeling device may include: aprocessor 1001 , such as a CPU, amemory 1005 , and acommunication bus 1002 . Among them, thecommunication bus 1002 is used to realize the connection communication between theprocessor 1001 and thememory 1005 . Thememory 1005 may be high-speed RAM memory, or may be non-volatile memory, such as disk memory. Optionally, thememory 1005 may also be a storage device independent of theaforementioned processor 1001 .

可选地，该联邦学习建模设备还可以包括矩形用户接口、网络接口、摄像头、RF(Radio Frequency，射频)电路，传感器、音频电路、WiFi模块等等。矩形用户接口可以包括显示屏(Display)、输入子模块比如键盘(Keyboard)，可选矩形用户接口还可以包括标准的有线接口、无线接口。网络接口可选的可以包括标准的有线接口、无线接口(如WI-FI接口)。Optionally, the federated learning modeling device may further include a rectangular user interface, a network interface, a camera, an RF (Radio Frequency, radio frequency) circuit, a sensor, an audio circuit, a WiFi module, and the like. The rectangular user interface may include a display screen (Display), an input sub-module such as a keyboard (Keyboard), and the optional rectangular user interface may also include a standard wired interface and a wireless interface. Optional network interfaces may include standard wired interfaces and wireless interfaces (eg, WI-FI interfaces).

本领域技术人员可以理解，图4中示出的联邦学习建模设备结构并不构成对联邦学习建模设备的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。Those skilled in the art can understand that the structure of the federated learning modeling device shown in FIG. 4 does not constitute a limitation on the federated learning modeling device, and may include more or less components than those shown in the figure, or combine some components, Or a different component arrangement.

如图4所示，作为一种计算机存储介质的存储器1005中可以包括操作系统、网络通信模块以及联邦学习建模方法程序。操作系统是管理和控制联邦学习建模设备硬件和软件资源的程序，支持联邦学习建模方法程序以及其它软件和/或程序的运行。网络通信模块用于实现存储器1005内部各组件之间的通信，以及与联邦学习建模方法系统中其它硬件和软件之间通信。As shown in FIG. 4 , thememory 1005 as a computer storage medium may include an operating system, a network communication module and a federated learning modeling method program. The operating system is a program that manages and controls the hardware and software resources of the federated learning modeling device, and supports the operation of the federated learning modeling method program and other software and/or programs. The network communication module is used to realize the communication between various components in thememory 1005, and communicate with other hardware and software in the federated learning modeling method system.

在图4所示的联邦学习建模设备中，处理器1001用于执行存储器1005中存储的联邦学习建模方法程序，实现上述任一项所述的联邦学习建模方法的步骤。In the federated learning modeling device shown in FIG. 4 , theprocessor 1001 is configured to execute the federated learning modeling method program stored in thememory 1005 to implement the steps of the federated learning modeling method described in any one of the above.

本申请联邦学习建模设备具体实施方式与上述联邦学习建模方法各实施例基本相同，在此不再赘述。The specific implementation manner of the federated learning modeling device of the present application is basically the same as the above-mentioned embodiments of the federated learning modeling method, and details are not repeated here.

本申请实施例还提供一种联邦学习建模装置，所述联邦学习建模装置应用于第一设备，所述联邦学习建模装置包括：An embodiment of the present application further provides a federated learning modeling apparatus, where the federated learning modeling apparatus is applied to the first device, and the federated learning modeling apparatus includes:

接收模块，用于接收各第二设备发送的加密模型参数和所述加密模型参数对应的验证参数；a receiving module, configured to receive the encryption model parameters sent by each second device and the verification parameters corresponding to the encryption model parameters;

零知识验证模块，用于基于各所述验证参数，分别对各所述加密模型参数进行零知识验证，以在各所述加密模型参数中确定虚假加密模型参数，获得零知识验证结果；A zero-knowledge verification module, configured to perform zero-knowledge verification on each of the encryption model parameters based on each of the verification parameters, so as to determine false encryption model parameters in each of the encryption model parameters, and obtain a zero-knowledge verification result;

协调模块，用于基于所述零知识验证结果和各所述加密模型参数，协调各所述第二设备进行联邦学习建模。A coordination module, configured to coordinate each of the second devices to perform federated learning modeling based on the zero-knowledge verification result and each of the encryption model parameters.

可选地，所述协调模块包括：Optionally, the coordination module includes:

剔除子模块，用于基于所述零知识验证结果，在各所述加密模型参数中剔除所述虚假加密模型参数，获得各可信模型参数；Elimination sub-module, for eliminating the false encryption model parameters from the encryption model parameters based on the zero-knowledge verification result, to obtain each trusted model parameter;

聚合子模块，用于对各所述可信模型参数进行聚合处理，获得聚合参数，并将所述聚合参数分别反馈至各所述第二设备，以供各所述第二设备更新各自的本地训练模型，直至所述本地训练模型达到预设训练结束条件。an aggregation sub-module, configured to perform aggregation processing on each of the trusted model parameters, obtain aggregation parameters, and feed back the aggregation parameters to each of the second devices, so that each of the second devices can update their local Train the model until the local training model reaches the preset training end condition.

可选地，所述零知识验证模块包括：Optionally, the zero-knowledge verification module includes:

计算子模块，用于分别计算各所述验证参数对应的第一零知识证明结果和第二零知识证明结果；a calculation submodule for calculating the first zero-knowledge proof result and the second zero-knowledge proof result corresponding to each of the verification parameters;

零知识验证子模块，用于基于各所述第一零知识证明结果和各所述第二零知识证明结果，分别验证各所述加密模型参数是否为虚假加密模型参数，获得零知识验证结果。The zero-knowledge verification sub-module is configured to verify whether each of the encryption model parameters is a fake encryption model parameter based on each of the first zero-knowledge proof results and each of the second zero-knowledge proof results, and obtain a zero-knowledge verification result.

可选地，所述计算子模块包括：Optionally, the calculation submodule includes:

合法性验证单元，用于基于预设验证挑战参数，对各所述加密模型参数进行合法性验证，获得各所述第一零知识验证结果；a legality verification unit, configured to perform legality verification on each of the encryption model parameters based on preset verification challenge parameters, and obtain each of the first zero-knowledge verification results;

加密单元，用于基于预设协调方公钥和各所述验证随机参数，对各所述验证模型参数进行加密处理，获得各所述第二零知识验证结果。An encryption unit, configured to perform encryption processing on each of the verification model parameters based on the preset coordinator public key and each of the verification random parameters to obtain each of the second zero-knowledge verification results.

可选地，所述合法性验证单元包括：Optionally, the legality verification unit includes:

第一幂操作子单元，用于分别将所述第一验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第一幂操作结果；a first exponentiation operation subunit, configured to perform an exponentiation operation on the first verification challenge parameter and each of the encryption model parameters, respectively, to obtain a first exponentiation operation result corresponding to each of the encryption model parameters;

第二幂操作子单元，用于分别将所述第二验证挑战参数与各所述加密模型参数进行幂操作，获得各所述加密模型参数对应的第二幂操作结果；The second exponentiation operation subunit is used to perform an exponentiation operation on the second verification challenge parameter and each of the encryption model parameters, respectively, to obtain a second exponentiation operation result corresponding to each of the encryption model parameters;

生成子单元，用于基于各所述第一幂操作结果和各所述第二幂操作结果，生成各所述第一零知识验证结果。A generating subunit is configured to generate each of the first zero-knowledge verification results based on each of the first power operation results and each of the second power operation results.

可选地，所述零知识验证子模块包括：Optionally, the zero-knowledge verification submodule includes:

对比单元，用于将各所述加密模型参数对应的所述第一零知识证明结果与所述第二零知识证明结果分别进行对比；a comparison unit, configured to compare the first zero-knowledge proof result corresponding to each of the encryption model parameters with the second zero-knowledge proof result;

第一判定单元，用于若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果不一致，则判定所述加密模型参数为所述虚假加密模型参数；a first determination unit, configured to determine that the encryption model parameter is the false encryption model parameter if the first zero-knowledge proof result corresponding to the encryption model parameter is inconsistent with the second zero-knowledge proof result;

第二判定单元，用于若所述加密模型参数对应的所述第一零知识证明结果和所述第二零知识证明结果一致，则判定所述加密模型参数不为所述虚假加密模型参数。A second determination unit, configured to determine that the encryption model parameter is not the fake encryption model parameter if the first zero-knowledge proof result corresponding to the encryption model parameter is consistent with the second zero-knowledge proof result.

本申请联邦学习建模装置的具体实施方式与上述联邦学习建模方法各实施例基本相同，在此不再赘述。The specific implementation of the federated learning modeling apparatus of the present application is basically the same as the above-mentioned embodiments of the federated learning modeling method, and details are not described herein again.

为实现所述目的，本实施例还提供一种联邦学习建模装置，所述联邦学习建模装置应用于第二设备，所述联邦学习建模装置包括：To achieve the purpose, this embodiment also provides a federated learning modeling apparatus, the federated learning modeling apparatus is applied to the second device, and the federated learning modeling apparatus includes:

加密模块，用于获取模型训练参数和第一验证随机参数，并基于所述第一验证随机参数和预设公钥，对所述模型训练参数进行加密处理，获得加密模型参数；an encryption module, configured to obtain model training parameters and first verification random parameters, and based on the first verification random parameters and a preset public key, perform encryption processing on the model training parameters to obtain encrypted model parameters;

生成模块，用于基于所述第一验证随机参数、所述模型训练参数和预设验证挑战参数，生成验证模型参数和第二验证随机参数；a generation module, configured to generate a verification model parameter and a second verification random parameter based on the first verification random parameter, the model training parameter and the preset verification challenge parameter;

发送模块，用于将所述加密模型参数、所述验证模型参数和所述第二验证随机参数发送至第一设备，以供所述第一设备进行零知识验证，获得零知识验证结果；a sending module, configured to send the encryption model parameter, the verification model parameter and the second verification random parameter to the first device, so that the first device can perform zero-knowledge verification to obtain a zero-knowledge verification result;

模型更新模块，用于接收所述第一设备基于所述零知识验证结果和所述加密模型参数反馈的聚合参数，并基于所述聚合参数，对所述模型训练参数对应的本地训练模型进行更新，直至所述本地训练模型达到预设训练结束条件。A model update module, configured to receive the aggregation parameters fed back by the first device based on the zero-knowledge verification result and the encryption model parameters, and update the local training model corresponding to the model training parameters based on the aggregation parameters , until the local training model reaches the preset training end condition.

可选地，所述加密模块包括：Optionally, the encryption module includes:

获取子模块，用于对所述模型训练参数对应的本地训练模型进行迭代训练，直至所述本地训练模型达到预设迭代次数阀值，获取所述本地训练模型的所述当前模型参数；an acquisition submodule, configured to perform iterative training on the local training model corresponding to the model training parameters, until the local training model reaches a preset number of iterations threshold, and acquire the current model parameters of the local training model;

生成子模块，用于获取所述本地训练模型的在前模型参数，并基于所述在前模型参数，生成所述辅助模型参数。A generating submodule is configured to acquire the previous model parameters of the local training model, and generate the auxiliary model parameters based on the previous model parameters.

本申请联邦学习建模装置的具体实施方式与上述联邦学习建模方法各实施例基本相同，在此不再赘述。The specific implementation of the federated learning modeling apparatus of the present application is basically the same as the above-mentioned embodiments of the federated learning modeling method, and details are not described herein again.

以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利处理范围内。The above are only the preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the contents of the description and drawings of the present application, or directly or indirectly applied in other related technical fields , are similarly included within the scope of patent processing of this application.

Claims (10)

1. The federated learning modeling method is applied to first equipment, and comprises the following steps:

receiving encryption model parameters sent by each second device and verification parameters corresponding to the encryption model parameters;

based on each verification parameter, respectively carrying out zero knowledge verification on each encryption model parameter so as to determine a false encryption model parameter in each encryption model parameter and obtain a zero knowledge verification result;

and coordinating each second device to carry out federated learning modeling based on the zero knowledge verification result and each encryption model parameter.

2. The federal learning modeling method as claimed in claim 1, wherein said step of coordinating each of said second devices for federal learning modeling based on said zero knowledge validation result and each of said cryptographic model parameters comprises:

based on the zero-knowledge verification result, eliminating the false encryption model parameters from the encryption model parameters to obtain the credible model parameters;

and aggregating the credible model parameters to obtain aggregated parameters, and feeding the aggregated parameters back to the second equipment respectively so that the second equipment can update the local training models thereof until the local training models reach the preset training ending condition.

3. The federal learning modeling method as claimed in claim 1, wherein the zero knowledge verification step of performing zero knowledge verification on each cryptographic model parameter based on each verification parameter, respectively, to determine a false cryptographic model parameter in each cryptographic model parameter, and obtaining a zero knowledge verification result comprises:

respectively calculating a first zero knowledge proof result and a second zero knowledge proof result corresponding to each verification parameter;

and respectively verifying whether each encryption model parameter is a false encryption model parameter or not based on each first zero knowledge proof result and each second zero knowledge proof result to obtain a zero knowledge verification result.

4. The federated learning modeling method of claim 3, wherein the verification parameters include verification model parameters and verification random parameters,

the step of calculating a first zero knowledge proof result and a second zero knowledge proof result corresponding to each verification parameter respectively includes:

carrying out validity verification on each encryption model parameter based on a preset verification challenge parameter to obtain each first zero knowledge verification result;

and encrypting each verification model parameter based on a preset coordinator public key and each verification random parameter to obtain each second zero knowledge verification result.

5. The federal learning modeling method of claim 4, wherein the preset verification challenge parameters include a first verification challenge parameter and a second verification challenge parameter;

the step of performing validity verification on each encryption model parameter based on a preset verification challenge parameter to obtain each first zero knowledge verification result comprises:

performing exponentiation operation on the first verification challenge parameter and each encryption model parameter respectively to obtain a first exponentiation operation result corresponding to each encryption model parameter;

performing exponentiation operation on the second verification challenge parameter and each encryption model parameter respectively to obtain a second exponentiation operation result corresponding to each encryption model parameter;

generating each of the first zero knowledge verification results based on each of the first exponentiation results and each of the second exponentiation results.

6. The federal learning modeling method as claimed in claim 3, wherein said step of verifying whether each of said cryptographic model parameters is a dummy cryptographic model parameter based on each of said first zero knowledge proof results and each of said second zero knowledge proof results, respectively, comprises:

comparing the first zero knowledge proof result and the second zero knowledge proof result corresponding to each encryption model parameter respectively;

if the first zero knowledge proof result and the second zero knowledge proof result corresponding to the encryption model parameter are not consistent, determining that the encryption model parameter is the false encryption model parameter;

and if the first zero knowledge proof result and the second zero knowledge proof result corresponding to the encryption model parameter are consistent, judging that the encryption model parameter is not the false encryption model parameter.

7. The federated learning modeling method is applied to second equipment, and comprises the following steps:

acquiring a model training parameter and a first verification random parameter, and encrypting the model training parameter based on the first verification random parameter and a preset public key to obtain an encrypted model parameter;

generating a verification model parameter and a second verification random parameter based on the first verification random parameter, the model training parameter and a preset verification challenge parameter;

sending the encryption model parameter, the verification model parameter and the second verification random parameter to first equipment so that the first equipment can carry out zero knowledge verification to obtain a zero knowledge verification result;

and receiving an aggregation parameter fed back by the first equipment based on the zero knowledge verification result and the encryption model parameter, and updating a local training model corresponding to the model training parameter based on the aggregation parameter until the local training model reaches a preset training end condition.

8. The federated learning modeling method of claim 7, wherein the model training parameters include current model parameters and auxiliary model parameters,

the step of obtaining model training parameters comprises:

performing iterative training on a local training model corresponding to the model training parameters until the local training model reaches a preset iteration threshold value, and acquiring the current model parameters of the local training model;

and acquiring the prior model parameters of the local training model, and generating the auxiliary model parameters based on the prior model parameters.

9. The federal learning modeling apparatus is characterized in that the federal learning modeling apparatus includes: a memory, a processor, and a program stored on the memory for implementing the federated learning modeling method,

the memory is used for storing a program for realizing the federal learning modeling method;

the processor is configured to execute a program implementing the federal learning modeling method to implement the steps of the federal learning modeling method as claimed in any of claims 1 to 6 or 7 to 8.

10. A readable storage medium having stored thereon a program for implementing a federal learning modeling method, the program being executed by a processor to implement the steps of the federal learning modeling method as claimed in any of claims 1 to 6 or 7 to 8.

Images