CN113642844A - Federal learning task management method and system - Google Patents
Federal learning task management method and system Download PDFInfo
- Publication number
- CN113642844A CN113642844A CN202110784437.9A CN202110784437A CN113642844A CN 113642844 A CN113642844 A CN 113642844A CN 202110784437 A CN202110784437 A CN 202110784437A CN 113642844 A CN113642844 A CN 113642844A
- Authority
- CN
- China
- Prior art keywords
- tenant
- information
- client
- learning task
- task
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000007726 management method Methods 0.000 title claims description 172
- 238000000034 method Methods 0.000 claims abstract description 28
- 230000000977 initiatory effect Effects 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 17
- 238000003860 storage Methods 0.000 claims description 10
- 239000003999 initiator Substances 0.000 claims description 6
- 230000006399 behavior Effects 0.000 abstract description 14
- 238000010586 diagram Methods 0.000 description 25
- 230000006870 function Effects 0.000 description 7
- 230000007246 mechanism Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000013524 data verification Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000007664 blowing Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Storage Device Security (AREA)
Abstract
The invention discloses a method and a system for managing a federated learning task, which are applied to the technical field of block chains, wherein the method comprises the following steps: the method comprises the steps that a management platform receives a federal learning task initiated by a client of a first tenant, wherein the federal learning task carries task information; the management platform initiates a federal learning task to the block chain; the method comprises the steps that a management platform receives federated learning task progress information and federated learning task result information fed back by a block chain, the federated learning task progress information is sent to the block chain when a client of a first tenant and a client of a second tenant perform joint modeling, and the joint modeling is started after the client of the second tenant receives a federated learning task initiated by the client of the first tenant. The invention can manage the federal learning task, is convenient for users to use and effectively restricts the behavior of the participants on the basis of realizing the safety and credibility of data.
Description
Technical Field
The invention relates to the technical field of block chains, in particular to a federal learning task management method and a system.
Background
In recent years, artificial intelligence can be called wind, fire and fire, a wave and a wave tide are raised, and AI gradually enters the aspect of people's life from face recognition to human chess players of alpha dog war and then to unmanned driving and generally applied precise marketing. Of course, partial over-blowing is not avoided, which leads to misunderstanding of the AI: AI is impossible. While tracking the AI, neglecting a little, the AI is supported by data and is a large amount of quality data.
In real life, except a few huge companies, most enterprises have the problems of small data volume and poor data quality, and the realization of artificial intelligence technology is not enough supported; meanwhile, the domestic and foreign supervision environment also gradually strengthens data protection, so that data freely flows on the premise of safety compliance, and becomes a great trend; data owned by business companies often has great potential value from both a user and enterprise perspective. Two companies and even departments between companies are concerned with the exchange of interests, and often these organizations do not provide for the aggregation of individual data with other companies, resulting in data that often appears in an isolated island even within the same company.
Based on the three points that the realization is not supported enough, the rough exchange is not allowed, and the value is not willing to be contributed, the data islands existing in large quantity at present are caused, and the privacy protection problem is solved, and the federal study is carried forward. Federal learning is a machine learning framework and can effectively help a plurality of organizations to carry out data use and machine learning modeling under the condition of meeting the requirements of user privacy protection, data safety and government regulations. Federal learning time is also directly referred to as federal modeling. Federal learning is essentially a distributed machine learning technique, or machine learning framework. The federal learning aims to realize common modeling and improve the effect of an AI model on the basis of ensuring the data privacy safety and legal compliance. Currently the federal learning technology framework is represented by FATE. FATE (Federated AI Technology Enabler) is an open source project initiated by the AI department of the micro-Bank and provides a reliable and safe computing framework for the federal learning ecosystem. The FATE project uses multi-party security computing (MPC) and Homomorphic Encryption (HE) technology to construct an underlying security computing protocol, so as to support security computing of different kinds of machine learning, including logistic regression, tree-based algorithm, deep learning, migration learning and the like.
Federal learning has attracted extensive attention as an emerging technology, and there are two main approaches to the prior art: training through a centralized system-centric scheduling model, or fully distributed peer-to-peer connections. However, the centralized system is not easily accepted by data providers, and has data security risks. Although the distributed system is more secure in data, the distributed system is difficult to deploy, complicated in configuration and high in user use threshold.
Disclosure of Invention
The embodiment of the invention provides a federated learning task management method, which is used for carrying out federated learning task management, is convenient for users to use and effectively restricts the behavior of participants on the basis of realizing safe and credible data, and comprises the following steps:
a management platform receives a federal learning task initiated by a client of a first tenant, wherein the federal learning task carries task information;
the management platform initiates the federal learning task to the block chain;
the method comprises the steps that a management platform receives federated learning task progress information and federated learning task result information fed back by a block chain, the federated learning task result information is sent to the block chain after a client of a first tenant and a client of a second tenant complete combined modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out combined modeling, the combined modeling is started after the client of the second tenant receives a federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates the federated learning task to the client of the corresponding second tenant according to task information after the block chain receives the federated learning task initiated by the management platform.
The embodiment of the invention provides a management platform, which is used for carrying out federal learning task management, is convenient for users to use and effectively restricts the behavior of participants on the basis of realizing data safety and credibility, and comprises the following components:
the task receiving module is used for receiving a federal learning task initiated by a client of a first tenant, wherein the federal learning task carries task information;
the task initiating module is used for initiating the federal learning task to the block chain;
the information receiving module is used for receiving federated learning task progress information and federated learning task result information fed back by the block chain by the management platform, the federated learning task result information is sent to the block chain after the client of the first tenant and the client of the second tenant complete the joint modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out the joint modeling, the joint modeling is started after the client of the second tenant receives the federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates the federated learning task to the client of the corresponding second tenant according to the task information after the block chain receives the federated learning task initiated by the management platform.
The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the federal learning task management method.
An embodiment of the present invention further provides a computer-readable storage medium, where a computer program for executing the federal learning task management method is stored in the computer-readable storage medium.
The embodiment of the invention receives a federal learning task initiated by a client of a first tenant through a management platform, wherein the federal learning task carries task information; the management platform initiates the federal learning task to the block chain; the method comprises the steps that a management platform receives federated learning task progress information and federated learning task result information fed back by a block chain, the federated learning task result information is sent to the block chain after a client of a first tenant and a client of a second tenant complete joint modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out joint modeling, the joint modeling is started after the client of the second tenant receives a federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates a federated learning task to the client of the corresponding second tenant according to the task information after the block chain receives the federated learning task initiated by the management platform. The embodiment of the invention can realize the safety and credibility of data through the federal learning of block chain scheduling, is convenient for users to use, can effectively restrict the behaviors of participants due to the data verification by using the block chain, and can judge the responsibility according to the data on the chain after disputes occur.
The embodiment of the invention provides a federated learning task management method, which is used for carrying out federated learning task management, is convenient for users to use and effectively restricts the behavior of participants on the basis of realizing safe and credible data, and comprises the following steps:
a client of a first tenant initiates a federal learning task to a management platform, wherein the federal learning task carries task information;
after a block chain receives a federal learning task initiated by a management platform, a client of a first tenant initiates the federal learning task to a client of a corresponding second tenant according to the task information, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the client of the first tenant;
after the client of the second tenant receives the federal learning task, the client of the first tenant and the client of the second tenant start to perform combined modeling and send progress information of the federal learning task to the block chain;
after the joint modeling is completed, the client of the first tenant and the client of the second tenant send the result information of the federal learning task to the block chain.
The embodiment of the invention provides a client for carrying out federal learning task management, which can be conveniently used by a user and effectively restrict the behavior of a participant on the basis of realizing data security and credibility, and the client comprises:
the system comprises a first task initiating module, a second task initiating module and a processing module, wherein the first task initiating module is used for initiating a Federal learning task to a management platform by a client of a first tenant, and the Federal learning task carries task information;
the second task initiating module is used for initiating a federal learning task to a corresponding second tenant client according to the task information after the block chain receives the federal learning task initiated by the management platform, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the first tenant client;
the combined modeling module is used for starting combined modeling by the client of the first tenant and the client of the second tenant and sending the progress information of the federated learning task to the block chain after the client of the second tenant receives the federated learning task;
and the information sending module is used for sending the federal learning task progress information to the block chain when the client of the first tenant and the client of the second tenant perform combined modeling, and sending the federal learning task result information to the block chain after the combined modeling is completed.
The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the federal learning task management method.
An embodiment of the present invention further provides a computer-readable storage medium, where a computer program for executing the federal learning task management method is stored in the computer-readable storage medium.
The embodiment of the invention initiates a federal learning task to a management platform through a client of a first tenant, wherein the federal learning task carries task information; after a block chain receives a federal learning task initiated by a management platform, a client of a first tenant initiates the federal learning task to a client of a corresponding second tenant according to the task information, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the client of the first tenant; after a client of a second tenant receives a federated learning task, the client of the first tenant and the client of the second tenant start to perform joint modeling and send federated learning task progress information to a block chain; after the joint modeling is completed, the client of the first tenant and the client of the second tenant send the result information of the federal learning task to the block chain. The embodiment of the invention can realize data security and credibility and is convenient for users to use through the federated learning of block chain scheduling, and can effectively restrict the behavior of the participants because the block chain is used for data verification, and after disputes occur, the responsibility judgment is carried out according to the data on the chain.
The embodiment of the invention provides a federated learning task management system, which is used for carrying out federated learning task management, is convenient for users to use and effectively restricts the behavior of participants on the basis of realizing safe and credible data, and comprises the following steps: the management platform and the client.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts. In the drawings:
FIG. 1 is a schematic diagram of a federated learning task management method in an embodiment of the present invention;
FIG. 2 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIG. 7 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIG. 8 is a diagram illustrating another federated learning task management method according to an embodiment of the present invention;
FIGS. 9-10 are schematic diagrams of a federated learning task management method in an embodiment of the present invention;
FIG. 11 is a diagram of a federated learning task management device in an embodiment of the present invention;
FIG. 12 is a diagram of a management platform architecture in an embodiment of the present invention;
FIG. 13 is a diagram illustrating a client architecture in accordance with an embodiment of the present invention;
fig. 14 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention and not to limit the present invention.
As mentioned above, there are two main ways in the prior art, centralized management scheduling system or distributed peer-to-peer system: 1. the centralized system is not easy to be accepted by data providers, and potential safety hazards exist. The schematic diagram is as follows, the computing nodes are directly scheduled by the management center, and the data providers (computing nodes) have data leakage risks and are uncontrollable. 2. Although the distributed peer-to-peer system has safer data, the identity data are respectively managed, so that the use is inconvenient, and the responsibility determination is difficult once disputes occur in the federal modeling study of two parties or multiple parties.
In order to perform federal learning task management, which can be conveniently used by a user and effectively restrict the behavior of a participant on the basis of realizing safe and credible data, an embodiment of the present invention provides a method for managing a federal learning task, as shown in fig. 1, the method may include:
102, the management platform initiates the federal learning task to a block chain;
103, the management platform receives federated learning task progress information and federated learning task result information fed back by the block chain, the federated learning task result information is sent to the block chain after the client of the first tenant and the client of the second tenant complete the joint modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out the joint modeling, the joint modeling is started after the client of the second tenant receives the federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates the federated learning task to the client of the corresponding second tenant according to the task information after the block chain receives the federated learning task initiated by the management platform.
As shown in fig. 1, in the embodiment of the present invention, a federal learning task initiated by a client of a first tenant is received through a management platform, where the federal learning task carries task information; the management platform initiates the federal learning task to the blockchain; the method comprises the steps that a management platform receives federated learning task progress information and federated learning task result information fed back by a block chain, the federated learning task result information is sent to the block chain after a client of a first tenant and a client of a second tenant complete combined modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out combined modeling, the combined modeling is started after the client of the second tenant receives a federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates the federated learning task to the client of the corresponding second tenant according to task information after the block chain receives the federated learning task initiated by the management platform. The embodiment of the invention can realize the safety and credibility of data through the federal learning of block chain scheduling, is convenient for users to use, can effectively restrict the behaviors of participants due to the data verification by using the block chain, and can judge the responsibility according to the data on the chain after disputes occur.
In an embodiment, the task information includes: initiator information, participant information, data hash list information and task configuration information or any combination thereof.
In an embodiment, as shown in fig. 2, the federal learning task management method further includes:
In this embodiment, the first registration information includes: identity information, public key information and compute node information of the first tenant.
In this embodiment, the blockchain is registered with first local data of a client of a first tenant, where the first local data includes: and the data hash information, the description information and the attribution information of the first tenant are one or any combination.
In an embodiment, as shown in fig. 3, the federal learning task management method further includes:
In this embodiment, the second registration information includes: identity information, public key information and compute node information of the second tenant.
In this embodiment, the blockchain is registered with second local data of a client of a second tenant, where the second local data includes: and the data hash information, the description information and the attribution information of the second tenant or any combination thereof.
In order to perform federal learning task management, which can be conveniently used by a user and effectively restrict the behavior of a participant on the basis of realizing data security and credibility, an embodiment of the present invention provides another federal learning task management method, as shown in fig. 4, which may include:
and step 404, after the joint modeling is completed, the client of the first tenant and the client of the second tenant send federated learning task result information to the blockchain.
As shown in fig. 4, in the embodiment of the present invention, a client of a first tenant initiates a federal learning task to a management platform, where the federal learning task carries task information; after a block chain receives a federal learning task initiated by a management platform, a client of a first tenant initiates the federal learning task to a client of a corresponding second tenant according to the task information, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the client of the first tenant; after the client of the second tenant receives the federal learning task, the client of the first tenant and the client of the second tenant start to perform combined modeling and send progress information of the federal learning task to the block chain; after the joint modeling is completed, the client of the first tenant and the client of the second tenant send the result information of the federal learning task to the block chain. The embodiment of the invention can realize the safety and credibility of data through the federal learning of block chain scheduling, is convenient for users to use, can effectively restrict the behavior of the participants because of using the block chain to store data certificates, and can perform duty judgment according to the data on the chain after disputes occur.
In an embodiment, the task information includes: initiator information, participant information, data hash list information and task configuration information or any combination thereof.
In an embodiment, as shown in fig. 5, the federal learning task management method further includes:
In an embodiment, as shown in fig. 6, the federal learning task management method in fig. 5 further includes:
In an embodiment, as shown in fig. 7, the federal learning task management method further includes:
In an embodiment, as shown in fig. 8, the federal learning task management method in fig. 7 further includes:
A specific embodiment is given below to illustrate a specific application of the federal learning task management method in an embodiment of the present invention. In this embodiment, as shown in fig. 9 to 10, federal learning task management is performed according to the following steps:
1. identity registration: each tenant accesses the management platform through a browser and fills in registration information, wherein the registration information comprises identity information, public key information and computing node information required by applying for a block chain account.
2. The management platform initiates an identity information binding request to the block chain, wherein the identity information binding request comprises identity information, public key information and computing node information.
3. Data registration: each tenant can register local data on the block chain, and the main content includes data hash information, description information, attribution information and the like.
4. And the management platform receives the block chain notification and synchronously obtains the data directory.
5. And (3) task initiation: the tenant of the mechanism 1 logs in the management platform through the browser, and selects the data of the mechanism 2 for combined modeling. This task will also be performed by the blockchain. The task content includes initiator information, participant information, data hash list information, and configuration information required by the task. After the mechanism 1 synchronous block chain is informed, the tenant of the mechanism is informed of initiating a task, and the modeling task is initiated to a specified party through loading configuration. The compute nodes of authority 2 passively accept tasks and actively participate. The modeling process is generally long in time and interactive. And the computing nodes of the mechanism 1 and the mechanism 2 can actively report the task progress to the block chain until the modeling is finished. The management platform can obtain task progress and results in real time through block chain notification. The tenant (task initiator) of organization 1 may view through the management platform. The whole process computing node and the management platform are interacted through the block chain, physical connection cannot exist, the risk of data leakage is avoided, and the data security problem is solved.
The embodiment of the invention realizes corresponding service functions of identity management, data management (data market) and task management through a block chain intelligent contract; the management platform is responsible for visualization configuration and usage. The method can comprise basic functions of task initiation, task state display, task result display and the like. User information, data markets, task states and task results are all derived from the blockchain. And the computing node only interacts with the block chain, registers data on the block chain, receives task scheduling of the block chain and feeds back a task state to the block chain. And the block chain is used as a medium, and the functions of identity management, data management and task management are realized on a management platform. And the trust is enhanced for the nodes of the federal learning participants, and the implementation landing is convenient. Data registration and task management are performed through a block chain, the whole information is transparent and credible, and the computing node can directly respond to related events. The identity, the data and the task all store corresponding relations in the block chain, and the whole process can be checked and traced.
Based on the same inventive concept, the embodiment of the present invention further provides a management platform, as described in the following embodiments. Because the principle of solving the problems is similar to the federal learning task management method, the implementation of the management platform can be referred to the implementation of the method, and repeated details are not repeated.
Fig. 11 is a structural diagram of a federal learning task management device in an embodiment of the present invention, and as shown in fig. 11, the management platform 1100 includes:
the task receiving module 1101 is configured to receive a federal learning task initiated by a client of a first tenant, where the federal learning task carries task information;
a task initiating module 1102, configured to initiate the federal learning task to a blockchain;
the information receiving module 1103 is configured to receive, by the management platform, federal learning task progress information and federal learning task result information fed back by the block chain, where the federal learning task result information is sent to the block chain after the client of the first tenant and the client of the second tenant complete joint modeling, the federal learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant perform joint modeling, the joint modeling is started after the client of the second tenant receives the federal learning task initiated by the client of the first tenant, and the client of the first tenant initiates a federal learning task to the client of the corresponding second tenant according to the task information after the block chain receives the federal learning task initiated by the management platform.
In summary, in the embodiment of the present invention, a management platform receives a federal learning task initiated by a client of a first tenant, where the federal learning task carries task information; the management platform initiates the federal learning task to the block chain; the method comprises the steps that a management platform receives federated learning task progress information and federated learning task result information fed back by a block chain, the federated learning task result information is sent to the block chain after a client of a first tenant and a client of a second tenant complete joint modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out joint modeling, the joint modeling is started after the client of the second tenant receives a federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates a federated learning task to the client of the corresponding second tenant according to task information after the block chain receives the federated learning task initiated by the management platform. The embodiment of the invention can realize data safety and credibility through the federal learning of block chain scheduling, is convenient for users to use, can effectively restrict the behavior of the participating party because the block chain is used for data storage, and can judge the responsibility according to the data on the chain after disputes occur.
Based on the same inventive concept, the embodiment of the present invention further provides a client, as described in the following embodiments. Because the principle of solving the problems is similar to the federal learning task management method, the implementation of the client can be referred to the implementation of the method, and repeated details are not repeated.
Fig. 12 is a structural diagram of a client according to an embodiment of the present invention, and as shown in fig. 12, the client 1200 includes:
a first task initiating module 1201, configured to initiate a federal learning task to a management platform by a client of a first tenant, where the federal learning task carries task information;
the second task initiating module 1202 is configured to initiate a federal learning task to a client of a corresponding second tenant according to the task information after the client of the first tenant receives the federal learning task initiated by the management platform in the block chain, where the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the client of the first tenant;
the joint modeling module 1203 is configured to, after the client of the second tenant receives the federal learning task, start joint modeling by the client of the first tenant and the client of the second tenant, and send progress information of the federal learning task to the blockchain;
the information sending module 1204 is configured to send federal learning task progress information to the blockchain when the client of the first tenant and the client of the second tenant perform joint modeling, and send federal learning task result information to the blockchain after the joint modeling is completed.
In summary, in the embodiment of the present invention, a client of a first tenant initiates a federal learning task to a management platform, where the federal learning task carries task information; after a block chain receives a federal learning task initiated by a management platform, a client of a first tenant initiates the federal learning task to a client of a corresponding second tenant according to the task information, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the client of the first tenant; after the client of the second tenant receives the federal learning task, the client of the first tenant and the client of the second tenant start to perform joint modeling and send progress information of the federal learning task to the block chain; after the joint modeling is completed, the client of the first tenant and the client of the second tenant send the result information of the federal learning task to the block chain. The embodiment of the invention can realize the safety and credibility of data through the federal learning of block chain scheduling, is convenient for users to use, can effectively restrict the behavior of the participating party because the block chain is used for data storage, and can judge the responsibility according to the data on the chain after disputes occur.
Based on the same inventive concept, the embodiment of the present invention further provides a federated learning task management system, as described in the following embodiments. Because the principle of solving the problems is similar to the federal learning task management method, the implementation of the system can be referred to the implementation of the method, and repeated details are not repeated.
Fig. 13 is a structural diagram of a federal learning task management system in an embodiment of the present invention, and as shown in fig. 13, the federal learning task management system includes: management platform 1100 and client 1200.
Based on the aforementioned inventive concept, as shown in fig. 14, the present invention further provides a computer apparatus 1400, which includes a memory 1410, a processor 1420, and a computer program 1430 stored in the memory 1410 and operable on the processor 1420, wherein the processor 1420 implements the aforementioned federal learning task management method when executing the computer program 1430.
Based on the foregoing inventive concept, the present invention proposes a computer-readable storage medium storing a computer program which, when executed by a processor, implements the foregoing federal learning task management method.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (19)
1. A method for managing a federated learning task is characterized by comprising the following steps:
the method comprises the steps that a management platform receives a federal learning task initiated by a client of a first tenant, wherein the federal learning task carries task information;
the management platform initiates the federal learning task to the block chain;
the method comprises the steps that a management platform receives federated learning task progress information and federated learning task result information fed back by a block chain, the federated learning task result information is sent to the block chain after a client of a first tenant and a client of a second tenant complete joint modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out joint modeling, the joint modeling is started after the client of the second tenant receives a federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates the federated learning task to the corresponding client of the second tenant according to the task information after the block chain receives the federated learning task initiated by the management platform.
2. The federal learning task management method of claim 1, wherein the task information includes: initiator information, participant information, data hash list information and task configuration information or any combination thereof.
3. The federal learning task management method of claim 1, further comprising:
the method comprises the steps that a management platform receives a first registration block chain account request sent by a client of a first tenant;
the management platform sends information to be filled to a client of a first tenant according to the first registration block chain account request;
the management platform receives first registration information fed back by a client of a first tenant according to the information to be filled;
and the management platform initiates a first identity information binding request to the block chain, wherein the first identity information binding request carries the first registration information.
4. The federal learning task management method of claim 3, wherein the first registration information includes: identity information, public key information and compute node information of the first tenant.
5. The federated learning task management method of claim 3, wherein the blockchain is registered with first local data of a client of a first tenant, the first local data comprising: and the data hash information, the description information and the attribution information of the first tenant are one or any combination.
6. The federal learning task management method of claim 1, further comprising:
the management platform receives a second registration block chain account request sent by a client of a second tenant;
the management platform sends information to be filled to a client of a second tenant according to the second registration block chain account request;
the management platform receives second registration information fed back by the client of the second tenant according to the information to be filled;
and the management platform initiates a second identity information binding request to the block chain, wherein the second identity information binding request carries the second registration information.
7. The federal learning task management method of claim 6, wherein the second registration information includes: identity information, public key information and compute node information of the second tenant.
8. The federated learning task management method of claim 6, wherein the blockchain is registered with second local data for clients of a second tenant, the second local data comprising: and the data hash information, the description information and the attribution information of the second tenant or any combination thereof.
9. A method for managing a federated learning task is characterized by comprising the following steps:
a client of a first tenant initiates a federal learning task to a management platform, wherein the federal learning task carries task information;
after a block chain receives a federal learning task initiated by a management platform, a client of a first tenant initiates the federal learning task to a client of a corresponding second tenant according to the task information, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the client of the first tenant;
after the client of the second tenant receives the federal learning task, the client of the first tenant and the client of the second tenant start to perform joint modeling and send progress information of the federal learning task to the block chain;
after the joint modeling is completed, the client of the first tenant and the client of the second tenant send the result information of the federal learning task to the block chain.
10. The federal learning task management method of claim 9, wherein the task information includes: initiator information, participant information, data hash list information and task configuration information or any combination thereof.
11. The federal learning task management method of claim 9, further comprising:
a client of a first tenant sends a first registration block chain account request to a management platform;
a client of a first tenant receives information to be filled, which is sent by a management platform according to the first registration block chain account request;
the client of the first tenant sends first registration information corresponding to information to be filled to the management platform, wherein the first registration information is used for being carried in a first identity information binding request initiated by the management platform to the block chain.
12. The federal learning task management method of claim 11, further comprising:
a client of a first tenant registers first local data in a blockchain, the first local data comprising: and the data hash information, the description information and the attribution information of the first tenant are one or any combination.
13. The federal learning task management method of claim 9, further comprising:
a client of a second tenant sends a second registration block chain account request to the management platform;
a client of a second tenant receives information to be filled, which is sent by a management platform according to the request of the second registration block chain account;
and the client of the second tenant sends second registration information corresponding to the information to be filled to the management platform, wherein the second registration information is carried in a second identity information binding request initiated by the management platform to the block chain.
14. The federal learning task management method of claim 13, further comprising:
a client of a second tenant registers second local data in a blockchain, the second local data comprising: and the data hash information, the description information and the attribution information of the second tenant or any combination thereof.
15. A management platform, comprising:
the task receiving module is used for receiving a federal learning task initiated by a client of a first tenant, wherein the federal learning task carries task information;
the task initiating module is used for initiating the federal learning task to the block chain;
the information receiving module is used for receiving federated learning task progress information and federated learning task result information fed back by the block chain by the management platform, the federated learning task result information is sent to the block chain after the client of the first tenant and the client of the second tenant complete joint modeling, the federated learning task progress information is sent to the block chain when the client of the first tenant and the client of the second tenant carry out joint modeling, the joint modeling is started after the client of the second tenant receives the federated learning task initiated by the client of the first tenant, and the client of the first tenant initiates the federated learning task to the client of the corresponding second tenant according to the task information after the block chain receives the federated learning task initiated by the management platform.
16. A client, comprising:
the system comprises a first task initiating module, a second task initiating module and a management module, wherein the first task initiating module is used for initiating a federated learning task to a management platform by a client of a first tenant, and the federated learning task carries task information;
the second task initiating module is used for initiating a federal learning task to a corresponding second tenant client according to the task information after the block chain receives the federal learning task initiated by the management platform, and the management platform initiates the federal learning task to the block chain after receiving the federal learning task initiated by the first tenant client;
the combined modeling module is used for starting combined modeling by the client of the first tenant and the client of the second tenant and sending the progress information of the federal learning task to the block chain after the client of the second tenant receives the federal learning task;
and the information sending module is used for sending the federal learning task progress information to the block chain when the client of the first tenant and the client of the second tenant perform combined modeling, and sending the federal learning task result information to the block chain after the combined modeling is completed.
17. The utility model provides a bang study task management system which characterized in that includes: a management platform as claimed in claim 15 and a client as claimed in claim 16.
18. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 14 when executing the computer program.
19. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the method of any one of claims 1 to 14.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110784437.9A CN113642844A (en) | 2021-07-12 | 2021-07-12 | Federal learning task management method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110784437.9A CN113642844A (en) | 2021-07-12 | 2021-07-12 | Federal learning task management method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113642844A true CN113642844A (en) | 2021-11-12 |
Family
ID=78417137
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110784437.9A Pending CN113642844A (en) | 2021-07-12 | 2021-07-12 | Federal learning task management method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113642844A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114219098A (en) * | 2021-12-03 | 2022-03-22 | 华融融通（北京）科技有限公司 | Federal learning system based on parameter server |
| CN115865705A (en) * | 2022-11-29 | 2023-03-28 | 广发银行股份有限公司 | Multi-party federal learning system, method, equipment and medium based on supervision mechanism |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111768008A (en) * | 2020-06-30 | 2020-10-13 | 平安科技（深圳）有限公司 | Federal learning method, device, equipment and storage medium |
| CN111950739A (en) * | 2020-08-13 | 2020-11-17 | 深圳前海微众银行股份有限公司 | Data processing method, device, equipment and medium based on block chain |
| CN112039848A (en) * | 2020-08-05 | 2020-12-04 | 北京链飞未来科技有限公司 | Web authentication method, system and device based on block chain public key digital signature |
| CN112328617A (en) * | 2020-11-19 | 2021-02-05 | 杭州趣链科技有限公司 | Learning mode parameter updating method for longitudinal federal learning and electronic device |
| CN112434818A (en) * | 2020-11-19 | 2021-03-02 | 脸萌有限公司 | Model construction method, device, medium and electronic equipment |
-
2021
- 2021-07-12 CN CN202110784437.9A patent/CN113642844A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111768008A (en) * | 2020-06-30 | 2020-10-13 | 平安科技（深圳）有限公司 | Federal learning method, device, equipment and storage medium |
| CN112039848A (en) * | 2020-08-05 | 2020-12-04 | 北京链飞未来科技有限公司 | Web authentication method, system and device based on block chain public key digital signature |
| CN111950739A (en) * | 2020-08-13 | 2020-11-17 | 深圳前海微众银行股份有限公司 | Data processing method, device, equipment and medium based on block chain |
| CN112328617A (en) * | 2020-11-19 | 2021-02-05 | 杭州趣链科技有限公司 | Learning mode parameter updating method for longitudinal federal learning and electronic device |
| CN112434818A (en) * | 2020-11-19 | 2021-03-02 | 脸萌有限公司 | Model construction method, device, medium and electronic equipment |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114219098A (en) * | 2021-12-03 | 2022-03-22 | 华融融通（北京）科技有限公司 | Federal learning system based on parameter server |
| CN114219098B (en) * | 2021-12-03 | 2024-06-11 | 华融融通（北京）科技有限公司 | Federal learning system based on parameter server |
| CN115865705A (en) * | 2022-11-29 | 2023-03-28 | 广发银行股份有限公司 | Multi-party federal learning system, method, equipment and medium based on supervision mechanism |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108876599B (en) | Poverty relief loan management system | |
| CN106843774A (en) | A kind of mass-rent construction method of the intelligent contract based on block chain | |
| Sun et al. | A trust management model to enhance security of cloud computing environments | |
| CN113965329A (en) | Cross-chain interoperation system, method, medium and data processing terminal | |
| CN109040271A (en) | A kind of network equipment completeness protection method under distributed environment | |
| CN110826903A (en) | Enterprise credibility assessment method based on block chain | |
| CN113642844A (en) | Federal learning task management method and system | |
| CN113065775A (en) | Bidding system based on block chain | |
| WO2020228531A1 (en) | Consortium blockchain governance method and apparatus, computer device and storage medium | |
| CN114579943A (en) | Employee digital identity management system and method based on block chain | |
| Zhang et al. | SABlockFL: a blockchain-based smart agent system architecture and its application in federated learning | |
| CN110197433A (en) | Numeric value transfer, apparatus and system | |
| CN113129149A (en) | Transaction risk identification method and device based on block chain and safe multi-party calculation | |
| Zhang et al. | Information diffusion at workplace | |
| CN112288213A (en) | High-reliability demand response system and method based on block chain | |
| CN113987080A (en) | Block chain excitation method and device based on reputation consensus and related products | |
| CN112115204A (en) | Multi-party security calculation implementation method based on block chain technology | |
| CN111415160B (en) | Bill-off business processing method and system based on blockchain and relevant nodes | |
| Li et al. | A blockchain-based educational digital assets management system | |
| CN116506227A (en) | Data processing method, device, computer equipment and storage medium | |
| CN116599641A (en) | Social method, system, equipment and storage medium based on blockchain | |
| CN115496601A (en) | Loan system and method for recording carbon-neutral carbon emission based on block chain | |
| CN115941242A (en) | Data polymorphic management method based on block chain | |
| CN113676494B (en) | Centralized data processing method and device | |
| Chen et al. | Security Threat Early Warning of Distance Education System Based on Blockchain |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20221024 Address after: 25 Financial Street, Xicheng District, Beijing 100033 Applicant after: CHINA CONSTRUCTION BANK Corp. Address before: 12 / F, 15 / F, No. 99, Yincheng Road, Shanghai pilot Free Trade Zone, 200120 Applicant before: Jianxin Financial Science and Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right |