CN112328610A - Information processing method and device based on block chain - Google Patents
Information processing method and device based on block chain Download PDFInfo
- Publication number
- CN112328610A CN112328610A CN202011061217.5A CN202011061217A CN112328610A CN 112328610 A CN112328610 A CN 112328610A CN 202011061217 A CN202011061217 A CN 202011061217A CN 112328610 A CN112328610 A CN 112328610A
- Authority
- CN
- China
- Prior art keywords
- block
- model
- precision
- reward
- parameter
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/23—Updating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses an information processing method and device based on a block chain. The method comprises the following steps: when the distribution condition of the block chain reward is met, acquiring the reward value of the block chain reward and the distribution proportion of the reward value among a block generating party, a first block verifying party and a second block verifying party, wherein the block generating party is a task participating party for generating a block updating artificial intelligence model, the first block verifying party is a task participating party for verifying a block and contributing local parameters, and the second block verifying party is a task participating party for verifying the block but not contributing the local parameters; calculating a first reward value, a second reward value and a third reward value according to the reward value and the distribution proportion; the stimulus value is dispensed using a dispensing channel provided by the blockchain. Therefore, all the participants can be stimulated to compete to generate a new block, local parameters are contributed to the community, and the model convergence speed is accelerated.
Description
Technical Field
The present application relates to the field of blockchains, and in particular, to a method and an apparatus for processing information based on blockchains.
Background
As is well known, artificial intelligence is also continuously promoted by information calculation and network communication at a rapid and rapid pace, but the artificial intelligence is obtained by continuously calculating and learning an artificial intelligence model based on a large amount of data, and the quality and scale of training data of the artificial intelligence model directly determine the excellent degree of the artificial intelligence.
Blockchains are an information technology emerging in recent years. In essence, the blockchain can be regarded as a shared database, and the data or information stored therein has the characteristics of being unforgeable, having no trace in the whole process, having traceability, having public transparency, having collective maintenance, and the like. Based on the characteristics, the block chain technology lays a solid 'trust' foundation, creates a reliable 'cooperation' mechanism and has wide application prospect.
As blockchain technology continues to mature, researchers in the neighborhood of artificial intelligence have also attempted to train out more powerful artificial intelligence through data generated by the blockchain market. Specifically, a block chain-based artificial intelligence model training scenario (BDML) is a block chain in which communities maintain an artificial intelligence model for solving a specific problem, and if a participant trains to obtain a better model, a new block is generated after voting by other participants, and each participant can perform collaborative training without disclosing own data.
This approach combines two powerful original resources: private machine learning, allowing training without revealing sensitive private data, and motivational mechanisms along blockchains that allow these systems to attract optimal data and models, making them more intelligent.
In the basic BDML setup, it is assumed that each participant is actively involved in model training, but this is not practical in the real world because model training incurs resource costs. Without a well-designed incentive mechanism, some knowledgeable participants may be reluctant to engage in model training. Therefore, how to design an effective incentive mechanism to encourage the participants to participate in the model training of BDML becomes a technical problem to be solved urgently by the DBML.
Disclosure of Invention
In view of the above problems, the present applicant inventively provides a method, apparatus, and storage medium for identifying a first state of a vehicle.
According to a first aspect of the embodiments of the present application, an information processing method based on a blockchain is applied to a task initiator that performs artificial intelligence model training by using the blockchain, and the method includes, when a distribution condition of the blockchain reward is satisfied, performing the following method to distribute the blockchain reward: acquiring an incentive value of block chain rewards and a distribution proportion of the incentive value among a block generator, a first block verifier and a second block verifier, wherein the block generator is a task participant generating a block update artificial intelligence model, the first block verifier is a task participant verifying a block and contributing local parameters, and the second block verifier is a task participant verifying the block but not contributing the local parameters; calculating a first reward value, a second reward value and a third reward value according to the reward value and the distribution proportion; the method comprises the steps of issuing a first reward value to a block generator by using an issuing channel provided by a block chain, issuing a second reward value to a first block verifier, and issuing a third reward value to a second block verifier.
According to an embodiment of the present application, before issuing the blockchain reward, the method further includes: establishing an artificial intelligence model and setting initialization information of the artificial intelligence model, wherein the initialization information comprises an initial model, initial parameters, initial model precision, a precision improvement threshold value and a method for calculating model precision; determining block generation conditions and block chain reward mechanisms of a block chain, wherein the block chain reward mechanisms comprise distribution conditions of block chain rewards, reward values of the block chain rewards and distribution proportions of the reward values among a block generation party, a first block verification party and a second block verification party; packing the artificial intelligence model, the initialization information of the artificial intelligence model, the block generation condition and the block chain reward mechanism into a first block of a block chain; the first chunk is uploaded to a chunk chain.
According to an embodiment of the present application, the triggering condition of the blockchain reward includes: when detecting that a block chain has new blocks generated; or when the account value of the income account is detected to be increased, wherein the income account is used for storing the transaction record of the income obtained by the artificial intelligence model.
According to a second aspect of the embodiments of the present application, an information processing method based on a block chain is applied to the block generator, and the method includes: acquiring a latest model, a latest parameter and a latest precision of the artificial intelligence model from the block chain as a first model, a first parameter and a first model precision; updating or not updating the first model as required to obtain a second model, and training the second model by using own data to obtain a second parameter; according to a method for calculating model precision acquired in advance, calculating second model precision corresponding to a second model and a second parameter and a precision improvement value obtained by comparing the second model precision with the first model precision; and judging whether the precision improvement value is greater than or equal to a pre-acquired precision improvement threshold value, if so, broadcasting a second model and a second parameter to the block verifier to obtain a voting result of the block verifier, judging whether the pre-acquired block generation condition is met according to the voting result, if so, generating a new block, and acquiring the reward issued to the block generator when the issuing condition of the block chain reward is met.
According to an embodiment of the present application, the block verifier includes a first block verifier and a second block verifier, and accordingly, broadcasts the second model and the second parameter to the block verifier to obtain a voting result of the block verifier, further including: broadcasting the second model and the second parameters to the block verifier to obtain local parameters contributed by the first block verifier; combining the second parameter and a third parameter contributed by the first block verifier to obtain a fourth parameter; and broadcasting the second model and the fourth parameter to the block verifier to obtain a voting result of the block verifier.
According to an embodiment of the present application, before obtaining the latest model, the latest parameters, and the latest precision of the artificial intelligence model from the blockchain, the method further includes: a method for obtaining the accuracy of the calculation model, an accuracy improvement threshold value and a block generation condition.
According to a third aspect of the embodiments of the present application, an information processing method based on a block chain is applied to a first block verifier, and the method includes: acquiring a latest model, a latest parameter and a latest precision of the artificial intelligence model from the block chain as a first model, a first parameter and a first model precision; receiving a second model and second parameters broadcast by the block generator; training the second model by using the own data to obtain a third parameter; calculating the precision of a second model corresponding to the second model and a third parameter according to a method for calculating the precision of the model, which is acquired in advance; judging whether the precision of the second model is greater than that of the first model, if so, returning a favorable voting result and returning a third parameter; if the block generator can generate a new block, the reward issued to the first block verifier is acquired when the issuing condition of the block chain reward is met.
According to a fourth aspect of the embodiments of the present application, an information processing method based on a block chain is applied to a second block verifier, and the method includes: acquiring a latest model, a latest parameter and a latest precision of the artificial intelligence model from the block chain as a first model, a first parameter and a first model precision; receiving a second model and second parameters broadcast by the block generator; training the second model by using the own data to obtain a third parameter; calculating the precision of a second model corresponding to the second model and a third parameter according to a method for calculating the precision of the model, which is acquired in advance; judging whether the precision of the second model is greater than that of the first model, if so, returning a favorable voting result; if the block generator can generate a new block, the reward issued to the second block verifier is acquired when the issuing condition of the block chain reward is met.
According to a fifth aspect of the embodiments of the present application, an information processing apparatus based on a blockchain is applied to a task initiator that performs artificial intelligence model training by using the blockchain, and the apparatus includes: the system comprises an incentive mechanism obtaining module, a block generation party, a first block verification party and a second block verification party, wherein the incentive mechanism obtaining module is used for obtaining an incentive value of block chain rewards and a distribution proportion of the incentive value among the block generation party, the first block verification party and the second block verification party, the block generation party is a task participant for generating a block updating artificial intelligence model, the first block verification party is a task participant for verifying a block and contributing local parameters, and the second block verification party is a task participant for verifying the block but not contributing the local parameters; the reward value distribution module is used for calculating a first reward value, a second reward value and a third reward value according to the reward value and the distribution proportion; and the reward value issuing module is used for issuing a first reward value to the block generating party by using an issuing channel provided by the block chain, issuing a second reward value to the first block verifying party and issuing a third reward value to the second block verifying party.
According to an embodiment of the present application, the apparatus further includes: the model creating module is used for creating an artificial intelligence model and setting initialization information of the artificial intelligence model, wherein the initialization information comprises an initial model, initial parameters, initial model precision, a precision improvement threshold value and a method for calculating the model precision; the block attribute determining module is used for determining block generation conditions and a block chain reward mechanism of the block chain, wherein the block chain reward mechanism comprises distribution conditions of block chain rewards, reward values of the block chain rewards and distribution proportions of the reward values among the block generating party, the first block verifying party and the second block verifying party; the first block production module is used for packaging the artificial intelligence model, the initialization information of the artificial intelligence model, the block generation condition and the block chain reward mechanism into a first block of a block chain; a first block upload module to upload the first block to a block chain.
According to a sixth aspect of the embodiments of the present application, an information processing apparatus based on a block chain, applied to a block generator, includes: the block acquisition module is used for acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as the first model, the first parameter and the first model precision; the model training module is used for updating or not updating the first model according to needs to obtain a second model, and training the second model by using own data to obtain a second parameter; the precision calculation module is used for calculating second model precision corresponding to the second model and the second parameter and a precision improvement value obtained by comparing the second model precision with the first model precision according to a pre-obtained method for calculating the model precision; and the block generation module is used for judging whether the precision improvement value is greater than or equal to a pre-acquired precision improvement threshold value, if so, broadcasting a second model and a second parameter to the block verification party to obtain a voting result of the block verification party, judging whether a pre-acquired block generation condition is met according to the voting result, if so, generating a new block, and acquiring the reward issued to the block generation party when the issuing condition of the block chain reward is met.
According to an embodiment of the present application, the block verifier includes a first block verifier and a second block verifier, and accordingly, the block generation module includes: the broadcasting submodule is used for broadcasting the second model and the second parameters to the block verifier to obtain the local parameters contributed by the first block verifier; the parameter obtaining submodule is used for combining the second parameter and a third parameter contributed by the first block verifier to obtain a fourth parameter; the broadcasting sub-module is further used for broadcasting the second model and the fourth parameter to the block verifier to obtain a voting result of the block verifier.
According to an embodiment of the present application, the apparatus further includes: and the block generation information acquisition module is used for acquiring a method for calculating the model precision, a precision improvement threshold and a block generation condition.
According to a seventh aspect of the embodiments of the present application, an information processing apparatus based on a block chain, applied to a first block verifier, includes: the block acquisition module is used for acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as the first model, the first parameter and the first model precision; a receiving module, configured to receive a second model and a second parameter broadcasted by the block generator; the model training module is used for training the second model by using own data to obtain a third parameter; the precision calculation module is used for calculating the precision of the second model corresponding to the second model and the third parameter according to a method for calculating the precision of the model acquired in advance; the training result feedback module is used for judging whether the precision of the second model is greater than that of the first model, and if so, returning a favorable voting result and returning a third parameter; and the reward acquisition module is used for acquiring the reward issued to the first block verification party when the block chain reward issuing condition is met if the block generator can generate a new block.
According to an eighth aspect of the embodiments of the present application, an information processing apparatus based on a block chain, applied to a second block verifier, includes: the block acquisition module is used for acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as the first model, the first parameter and the first model precision; a receiving module, configured to receive a second model and a second parameter broadcasted by the block generator; the model training module is used for training the second model by using own data to obtain a third parameter; the precision calculation module is used for calculating the precision of the second model corresponding to the second model and the third parameter according to a method for calculating the precision of the model acquired in advance; the training result feedback module is used for judging whether the precision of the second model is greater than that of the first model, and if so, returning a favorable voting result; and the reward acquisition module is used for acquiring the reward issued to the first block verification party when the block chain reward issuing condition is met if the block generator can generate a new block.
According to a ninth aspect of embodiments of the present application, an information processing system based on a block chain includes: the system comprises a task initiator for performing artificial intelligence model training by using a block chain, and an information processing method applied to the task initiator for performing artificial intelligence model training by using the block chain; a block generator for executing the information processing method applied to the block generator; the first block verifier is used for the information processing method applied to the first block verifier.
The embodiment of the application provides an information processing method and device based on a block chain, wherein the method comprises the following steps: firstly, a task initiator utilizing a block chain to train an artificial intelligence model gives model information such as an initial model, initial parameters, initial model precision, a precision improvement threshold value and a method for calculating the model precision, and reward information such as distribution conditions of block chain rewards, reward values of the block chain rewards and distribution proportions of the reward values among a block generator, a first block verifier and a second block verifier, and packages the information into a first block and uploads the first block to a block chain network, wherein the block generator is a task participant generating a block update artificial intelligence model, the first block verifier is a task participant verifying the block and contributing local parameters, and the second block verifier is a task participant verifying the block but not contributing local parameters; then, the participator can determine the used data volume, the computing resource, whether to integrate the parameters of other participators and other participation strategies according to the reward information; when the distribution condition of the block chain rewards is met, the following method is executed to distribute the block chain rewards: acquiring the reward value of the block chain reward and the distribution proportion of the reward value among the block generating party, the first block verifying party and the second block verifying party; calculating a first reward value, a second reward value and a third reward value according to the reward value and the distribution proportion; the method comprises the steps of issuing a first reward value to a block generator by using an issuing channel provided by a block chain, issuing a second reward value to a first block verifier, and issuing a third reward value to a second block verifier.
The incentive mechanism determines the allocation proportion among the block generator, the first block verifier and the second block verifier according to the roles of the participants and the contribution to model evolution when distributing incentives, and can more fully stimulate the participants to compete to generate new blocks, actively participate in new model verification and contribute parameters to communities as long as the allocation proportion is formulated reasonably, so that the convergence speed of the model is accelerated, and the artificial intelligence model training can be smoothly promoted without causing project stagnation due to low participation of the participants.
It is to be understood that the teachings of this application need not achieve all of the above-described benefits, but rather that specific embodiments may achieve specific technical results, and that other embodiments of this application may achieve benefits not mentioned above.
Drawings
The above and other objects, features and advantages of exemplary embodiments of the present application will become readily apparent from the following detailed description read in conjunction with the accompanying drawings. Several embodiments of the present application are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which:
in the drawings, the same or corresponding reference numerals indicate the same or corresponding parts.
Fig. 1 is a schematic view of an application scenario of an information processing method based on a block chain according to an embodiment of the present application;
fig. 2 is a schematic flowchart of an implementation process of an information processing method based on a block chain at a task initiator according to an embodiment of the present application;
fig. 3 is a schematic flow chart illustrating an implementation of the information processing method based on a block chain on a block generator according to the embodiment of the present application;
fig. 4 is a schematic flowchart illustrating an implementation process of the information processing method based on a block chain on a first block verifier according to the embodiment of the present application;
fig. 5 is a schematic flowchart illustrating an implementation process of the information processing method based on a block chain on a second block verifier according to the embodiment of the present application;
fig. 6 is a schematic structural diagram of a component of a block chain-based information processing apparatus at a task initiator according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of a block chain-based information processing apparatus on a block generating side according to an embodiment of the present application;
fig. 8 is a schematic structural diagram of a block chain-based information processing apparatus on a first block verifier according to an embodiment of the present application;
fig. 9 is a schematic structural diagram of a block chain-based information processing apparatus on a first block verifier according to an embodiment of the present application.
Detailed Description
In order to make the objects, features and advantages of the present application more obvious and understandable, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "a plurality" means two or more unless specifically limited otherwise.
Fig. 1 shows an application scenario of the information processing method based on the blockchain according to the embodiment of the present application. The application scene is an artificial intelligence model training scene based on a block chain.
As shown in fig. 1, a task initiator 20 for artificial intelligence model training may first create an artificial intelligence model, set an initial model, initial parameters, initial model precision, a method for precision improvement threshold and model precision calculation, and the like, and at the same time, the task initiator 20 may determine an incentive mechanism adopted for training the artificial intelligence model, including an issuing condition of a block chain reward, a reward value of a block chain reward, and a distribution ratio of the reward value among a block generator, a first block verifier and a second block verifier, and the like; then, the task initiator 20 packages the information into the first block (block 0) of the block chain 10 and uploads the first block onto the block chain 10.
A participant joining blockchain 10 for the first time downloads the information from block 0, and when the participant attempts to improve the artificial intelligence model, the participant pulls the latest model, the latest parameters, and the latest model precision from the latest block, say block 2, on blockchain 10, and then improves the model and/or trains the improved model using its own data to obtain another set of local parameters and local precision, and when the local precision exceeds the precision improvement threshold, the participant broadcasts the local parameters and the improved model to the other participants, wherein: some parties may participate in the verification using their own data and vote according to the verification result, such as the second block verifier 50; some participants will not only participate in validation and voting using their own data, but will also send the trained parameters to the participants, such as the first block validator 40; after receiving the parameters sent by the first verifier 40, the participant will synthesize the parameters with local parameters to generate new parameters, and then broadcast the new parameters to other participants again, and assuming that the voting result received after this broadcast meets the block generation condition, the participant will generate a new block (block 3) and upload the new block to the block chain 10. As such, the participant becomes a tile generator, namely tile generator 30 shown in fig. 1.
Assuming that in this application, the condition of reward distribution includes newly generating a block, at this time, the task initiator 20 distributes the corresponding reward value to the block generator 30, the first verifier 40 and the second verifier 50 through the distribution channel of the block chain 10 according to the preset reward value and distribution ratio. Thus, a new block generation process is completed. Because, the reward mechanism adopted by the embodiment of the application is based on: 1) whether the participator generates a new block or not, and 2) how much contribution is made to model improvement to distribute rewards, therefore, the participator can be better encouraged to train the artificial intelligence model by using own data, improve the existing model and contribute new parameters, thus the training progress of the model can be accelerated, the expected goal can be reached as soon as possible, and the training of the artificial intelligence model can be completed.
It should be noted that the above application is only one of application scenarios of the application embodiment, and is an exemplary illustration and is not a limitation on the application scenario of the application embodiment, and the application embodiment may also be applied to any other applicable application scenario.
The following describes in detail an information processing method and apparatus provided in the embodiments of the present application.
According to a first aspect of the embodiments of the present application, an information processing method based on a blockchain is applied to a task initiator that performs artificial intelligence model training by using the blockchain, as shown in fig. 2, the method includes, when an issue condition of the blockchain reward is satisfied, performing the following method to issue the blockchain reward: operation 210, obtaining an award value of the block chain award and an allocation ratio of the award value among a block generator, a first block verifier and a second block verifier, wherein the block generator is a task participant generating a block update artificial intelligence model, the first block verifier is a task participant verifying a block and contributing a local parameter, and the second block verifier is a task participant verifying a block but not contributing a local parameter; operation 220, calculating a first bonus value, a second bonus value and a third bonus value according to the bonus value and the distribution proportion; at operation 230, a first incentive value is issued to the tile generator using the issue channel provided by the blockchain, a second incentive value is issued to the first tile verifier, and a third incentive value is issued to the second tile verifier.
In operation 210, the blockchain reward is a reward given to the participant for stimulating the participant of the artificial intelligence model training task to train the artificial intelligence model using the own data and improve the artificial intelligence model according to the training result.
The bonus value refers primarily to a value or virtual currency value used to meter virtual assets, such as bitcoin, ethercoin, etc., which are well known to the art. In a blockchain implementation, these reward values may be mapped as an asset to a permit (Token).
The allocation proportions of the prize values among the block generator, the first block verifier and the second block verifier primarily refer to how much of the prize values in a given allocable prize value are available for allocation to the block generator, how much of the prize values are available for allocation to the first block verifier and how much of the prize values are available for allocation to the second block verifier. For example, assume that the preset allocation rule is: when a new block is produced, the producer of the block is given 50% of the prize value, all the first block verifiers are given 40% of the prize value, and more than the second block verifiers are given 10% of the prize value. In this case, the allocation ratio of the prize value among the block generator, the first block verifier and the second block verifier is 5: 4: 1, i.e. if the prize value is 100, the block generator can reach 50, all first block verifiers can reach 40, and all second block verifiers can reach 10.
In general, in order to encourage participants to improve the model as much as possible and to provide model parameters obtained after training using self-owned data, a lot of rewards are often given to the block generator, followed by the first block verification and finally the second block verifier. In an extreme case, the second block verifier may not even be assigned a prize value. Therefore, the effect that the incentive participants actively improve the model and contribute to the local parameters is achieved.
Updating artificial intelligence models, including but not limited to: adjustments to algorithms and/or adjustments to model parameters, etc.
And the verification block is used for verifying the updated artificial intelligence model broadcasted by other participants by using the own data, namely inputting the own data into the updated artificial intelligence model to obtain a corresponding output result, and calculating the model precision of the updated artificial intelligence model according to the output result and a specified model precision calculation method.
And contributing local parameters, namely training the updated artificial intelligence model broadcasted by other participants by using own data, and improving the parameters according to the training result to obtain new parameters, wherein the new parameters are the local parameters. The local parameters are then returned to the other participants broadcasting the updated artificial intelligence model to integrate the local parameters into the update of the artificial intelligence model. Thus, if other participants broadcasting updated artificial intelligence models successfully generate a block, the parameters of the artificial intelligence models in the block will include local parameters provided by the block verifier. In this manner, the tile verifier may also contribute local parameters.
It is obvious from the above analysis that the improvement of the artificial intelligence model is mainly attributed to the block generator and the first block verifier, so that each participant participating in the artificial intelligence model training is distinguished, and when the reward value is distributed, the distribution proportion of the block generator and the first block verifier is increased, so that each participant can be stimulated to actively participate in the module training and contribute to local parameters, thereby accelerating the progress of the artificial intelligence model and enabling the model to converge earlier to reach the expected target.
In operation 220 and operation 230, the first bonus value is a bonus value that the tile generator should receive, which is totally distributed to the tile generator; the second reward value is the reward value which all the first block verification parties can obtain, and is further distributed to each first block verification party according to the number of contributors or the number of contribution times; the third reward value is the reward value which all the second block verification parties can obtain, and is further distributed to each second block verification party according to the number of contributors or the number of contribution times;
in operation 230, dispensing the prize value generally refers to adding a credit record to the account of the corresponding participant and increasing the balance of the account by a corresponding amount, and accordingly, the address of the dispensing is the address of the account of the participant.
According to an embodiment of the present application, before issuing the blockchain reward, the method further includes: establishing an artificial intelligence model and setting initialization information of the artificial intelligence model, wherein the initialization information comprises an initial model, initial parameters, initial model precision, a precision improvement threshold value and a method for calculating model precision; determining block generation conditions and block chain reward mechanisms of a block chain, wherein the block chain reward mechanisms comprise distribution conditions of block chain rewards, reward values of the block chain rewards and distribution proportions of the reward values among a block generation party, a first block verification party and a second block verification party; packing the artificial intelligence model, the initialization information of the artificial intelligence model, the block generation condition and the block chain reward mechanism into a first block of a block chain; the first chunk is uploaded to a chunk chain.
The initial model and the initial parameters are the prototype of the artificial intelligence model, the information determines the basic algorithm of the artificial intelligence model, but the accuracy of the model can be continuously improved only by adjusting the algorithm and the parameters through the training of a large amount of data, so that the accuracy required by practical application is achieved.
The initial model precision is the first model precision standard of the artificial intelligence model, and the model precision in the newly generated block can only be higher than the initial model precision and cannot be lower than the initial model precision.
And the accuracy improvement threshold is mainly used for checking whether the updated artificial intelligence model has obvious improvement, and if the artificial intelligence model is only slightly improved, a new block is not suitable to be generated so as to avoid wasting the verification resources of other participants. Specifically, when a participant tries to improve the artificial intelligence model and wants to generate a new block, the accuracy of the updated artificial intelligence model is compared with the accuracy of the artificial intelligence model before updating to obtain an accuracy improvement value, if the accuracy improvement value is greater than the accuracy improvement threshold, the new block can be generated, otherwise, the artificial intelligence model needs to be continuously improved.
The method for calculating the model accuracy is used for providing a unified method for calculating the model accuracy so as to prevent different calculation methods from obtaining different results to influence the fairness of competition and the objectivity of the improvement degree of the model.
The block generation condition of the block chain generally relates to the precision improvement threshold mentioned above, and also relates to a condition that the block chain achieves consensus, for example, each block verifier votes according to the verification result, and determines whether the consensus is achieved or not according to the proportion of votes in the voting result, and a new block can be generated only when the consensus is achieved.
In general, the blockchain generation conditions and blockchain reward mechanisms are not easily changed once the blockchain is determined. Therefore, the tile generation conditions and the blockchain reward mechanism can be packaged into the first tile, and then, the participant who joins the blockchain for the first time can pull the tile generation conditions and the blockchain reward mechanism from the first tile to serve as rules and standards for guiding subsequent training tasks of the artificial intelligence model and new tile generation operation.
According to an embodiment of the present application, the triggering condition of the blockchain reward includes: when detecting that a block chain has new blocks generated; or when the account value of the income account is detected to be increased, wherein the income account is used for storing the transaction record of the income obtained by the artificial intelligence model.
In practical applications, any one of the above dispensing conditions may be selected, or two dispensing conditions may be used simultaneously as the dispensing conditions.
The detection means may be implemented by receiving a specific event through a certain port, or may be any other suitable method.
According to a second aspect of the embodiments of the present application, an information processing method based on a block chain is applied to the above block generator, as shown in fig. 3, the method includes: operation 310, obtaining the latest model, the latest parameter and the latest precision of the artificial intelligence model from the blockchain as the first model, the first parameter and the first model precision; operation 320, updating or not updating the first model as needed to obtain a second model, and training the second model by using own data to obtain a second parameter; operation 330, calculating a second model precision corresponding to the second model and the second parameter and a precision improvement value obtained by comparing the second model precision with the first model precision according to a pre-obtained method for calculating the model precision; operation 340 determines whether the precision improvement value is greater than or equal to a pre-obtained precision improvement threshold, if so, broadcasts a second model and a second parameter to the block verifier to obtain a voting result of the block verifier, determines whether a pre-obtained block generation condition is satisfied according to the voting result, if so, generates a new block, and obtains a reward issued to the block generator when a block chain reward issue condition is satisfied.
In operation 310, the latest model, the latest parameters, and the latest precision of the artificial intelligence model are typically the models, parameters, and precisions recorded in the last tile, so the latest model, the latest parameters, and the latest precision can be obtained by finding the last tile and pulling the models, parameters, and precisions therefrom.
In operation 320, the second parameter refers to a local parameter obtained by the block generator training the artificial model using the own data.
In operation 330, the pre-acquired method for calculating model accuracy task initiator sets the method for calculating model accuracy in the first block.
In operation 340, the pre-obtained precision increase threshold is the precision increase threshold set in the first block by the task initiator. The block generation condition obtained in advance is the block generation condition set in the first block by the task initiator.
According to an embodiment of the present application, the block verifier includes a first block verifier and a second block verifier, and accordingly, broadcasts the second model and the second parameter to the block verifier to obtain a voting result of the block verifier, further including: broadcasting the second model and the second parameters to the block verifier to obtain local parameters contributed by the first block verifier; combining the second parameter and a third parameter contributed by the first block verifier to obtain a fourth parameter; and broadcasting the second model and the fourth parameter to the block verifier to obtain a voting result of the block verifier.
In this embodiment, by further distinguishing the verification party that contributes to the local parameter from the verification party that does not contribute to the local parameter, more rewards can be allocated to the verification parties that contribute to the local parameter by increasing the allocation ratio to the first verification party, so that more participants are encouraged to contribute to the local parameter, the convergence progress is accelerated, and the training time is shortened.
According to an embodiment of the present application, before obtaining the latest model, the latest parameters, and the latest precision of the artificial intelligence model from the blockchain, the method further includes: a method for obtaining the accuracy of the calculation model, an accuracy improvement threshold value and a block generation condition.
Generally, the method for calculating model accuracy, the accuracy improvement threshold and the block generation condition are all pulled from the first block when the participant first joins the block chain, but this embodiment does not exclude obtaining these information from other ways, for example, from a certain agreed configuration information system, but whatever way is adopted, these information need to be consistent with the method for calculating model accuracy, the accuracy improvement threshold and the block generation condition set in the first block by the participant.
According to a third aspect of the embodiments of the present application, an information processing method based on a block chain, as shown in fig. 4, is applied to a first block verifier, and the method includes: operation 410, acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the blockchain as the first model, the first parameter and the first model precision; an operation 420 of receiving a second model and second parameters broadcast by the tile generator; operation 430, training the second model using the self-owned data to obtain a third parameter; operation 440, calculating a second model accuracy corresponding to the second model and the third parameter according to a pre-obtained method for calculating the model accuracy; operation 450, judging whether the second model precision is greater than the first model precision, if so, returning a favorable voting result and returning a third parameter; in operation 460, if the block generator can generate a new block, the reward issued to the first block verifier is obtained when the issue condition of the block chain reward is satisfied.
According to a fourth aspect of the embodiments of the present application, an information processing method based on a block chain is applied to a second block verifier, as shown in fig. 5, and the method includes: operation 510, obtaining the latest model, the latest parameter and the latest precision of the artificial intelligence model from the blockchain as the first model, the first parameter and the first model precision; operation 520, receiving a second model and second parameters broadcast by the block generator; operation 530, training the second model using the self-owned data to obtain a third parameter; operation 540, calculating a second model precision corresponding to the second model and the third parameter according to a pre-obtained method for calculating the model precision; operation 550, determining whether the second model precision is greater than the first model precision, and if so, returning a favorable voting result; in operation 560, if the block creator is able to create a new block, the reward issued to the second block verifier is obtained when the condition for issuing the blockchain reward is satisfied.
The other operations are similar to those of the method applied to the first block verifier except that the local parameter is not returned, and thus are not described again.
According to a fifth aspect of the embodiments of the present application, an information processing apparatus based on a blockchain is applied to a task initiator performing artificial intelligence model training by using the blockchain, as shown in fig. 6, the apparatus 60 includes: the incentive mechanism obtaining module 601 is configured to obtain an incentive value of a block chain incentive and a distribution ratio of the incentive value among a block generator, a first block verifier and a second block verifier, where the block generator is a task participant generating a block update artificial intelligence model, the first block verifier is a task participant verifying a block and contributing a local parameter, and the second block verifier is a task participant verifying a block but not contributing a local parameter; the reward value distribution module 602 is configured to calculate a first reward value, a second reward value, and a third reward value according to the reward value and the distribution proportion; the reward value issuing module 603 is configured to issue a first reward value to the block generator using the issue channel provided by the block chain, issue a second reward value to the first block verifier, and issue a third reward value to the second block verifier.
According to an embodiment of the present application, the apparatus 60 further includes: the model creating module is used for creating an artificial intelligence model and setting initialization information of the artificial intelligence model, wherein the initialization information comprises an initial model, initial parameters, initial model precision, a precision improvement threshold value and a method for calculating the model precision; the block attribute determining module is used for determining block generation conditions and a block chain reward mechanism of the block chain, wherein the block chain reward mechanism comprises distribution conditions of block chain rewards, reward values of the block chain rewards and distribution proportions of the reward values among the block generating party, the first block verifying party and the second block verifying party; the first block production module is used for packaging the artificial intelligence model, the initialization information of the artificial intelligence model, the block generation condition and the block chain reward mechanism into a first block of a block chain; a first block upload module to upload the first block to a block chain.
According to a sixth aspect of the embodiments of the present application, an information processing apparatus based on a block chain is applied to a block generator, as shown in fig. 7, the apparatus 70 includes: a block obtaining module 701, configured to obtain a latest model, a latest parameter, and a latest precision of the artificial intelligence model from the blockchain as a first model, a first parameter, and a first model precision; the model training module 702 is configured to update or not update the first model as needed to obtain a second model, and train the second model using own data to obtain a second parameter; the precision calculation module 703 is configured to calculate, according to a method for calculating model precision obtained in advance, a second model precision corresponding to the second model and the second parameter and a precision improvement value obtained by comparing the second model precision with the first model precision; the block generation module 704 is configured to determine whether the precision improvement value is greater than or equal to a precision improvement threshold value obtained in advance, if so, broadcast the second model and the second parameter to the block verifier to obtain a voting result of the block verifier, determine whether a block generation condition obtained in advance is satisfied according to the voting result, if so, generate a new block, and obtain a reward issued to the block generator when an issue condition of a block chain reward is satisfied.
According to an embodiment of the present application, the block verifier includes a first block verifier and a second block verifier, and accordingly, the block generating module 704 includes: the broadcasting submodule is used for broadcasting the second model and the second parameters to the block verifier to obtain the local parameters contributed by the first block verifier; the parameter obtaining submodule is used for combining the second parameter and a third parameter contributed by the first block verifier to obtain a fourth parameter; the broadcasting sub-module is further used for broadcasting the second model and the fourth parameter to the block verifier to obtain a voting result of the block verifier.
According to an embodiment of the present application, the apparatus 70 further includes: and the block generation information acquisition module is used for acquiring a method for calculating the model precision, a precision improvement threshold and a block generation condition.
According to a seventh aspect of the embodiment of the present application, an information processing apparatus based on a block chain is applied to a first block verifier, as shown in fig. 8, where the apparatus 80 includes: a block obtaining module 801, configured to obtain, from a block chain, a latest model, a latest parameter, and a latest precision of the artificial intelligence model as a first model, a first parameter, and a first model precision; a receiving module 802, configured to receive a second model and a second parameter broadcasted by the block generator; the model training module 803 is configured to train the second model using the own data to obtain a third parameter; the precision calculation module 804 is used for calculating the precision of the second model corresponding to the second model and the third parameter according to a method for calculating the precision of the model obtained in advance; a training result feedback module 805, configured to determine whether the second model precision is greater than the first model precision, and if so, return a favorable voting result and return a third parameter; the reward obtaining module 806 is configured to obtain a reward issued to the first block verifier when the condition for issuing the blockchain reward is satisfied if the block generator is able to generate a new block.
According to an eighth aspect of the embodiment of the present application, an information processing apparatus based on a block chain is applied to a second block verifier, as shown in fig. 9, where the apparatus 90 includes: a block obtaining module 901, configured to obtain, from a block chain, a latest model, a latest parameter, and a latest precision of the artificial intelligence model as a first model, a first parameter, and a first model precision; a receiving module 902, configured to receive a second model and a second parameter broadcasted by the block generator; the model training module 903 is used for training the second model by using own data to obtain a third parameter; a precision calculation module 904, configured to calculate, according to a method for calculating model precision obtained in advance, second model precision corresponding to the second model and the third parameter; the training result feedback module 905 judges whether the precision of the second model is greater than that of the first model, and returns a favorable voting result if the precision of the second model is greater than that of the first model; the reward acquiring module 906 is configured to acquire a reward issued to the first block verifier when the issue condition of the blockchain reward is satisfied if the block generator is able to generate a new block.
According to a ninth aspect of embodiments of the present application, an information processing system based on a block chain includes: the system comprises a task initiator for performing artificial intelligence model training by using a block chain, and an information processing method applied to the task initiator for performing artificial intelligence model training by using the block chain; a block generator for executing the information processing method applied to the block generator; the first block verifier is used for the information processing method applied to the first block verifier.
Here, it should be noted that: the above description of the embodiment of the information processing apparatus based on the blockchain and the above description of the embodiment of the information processing system based on the blockchain are similar to the description of the foregoing method embodiments, and have similar beneficial effects to the foregoing method embodiments, and therefore, the description thereof is omitted. For the technical details that have not been disclosed in the description of the embodiments of the blockchain-based information processing apparatus and the description of the embodiments of the blockchain-based information processing system, please refer to the description of the foregoing method embodiments of the present application for understanding, and therefore, for brevity, no further description is provided.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described device embodiments are merely illustrative, for example, the division of a unit is only one logical function division, and there may be other division ways in actual implementation, such as: multiple units or components may be combined, or may be integrated into another device, or some features may be omitted, or not implemented. In addition, the coupling, direct coupling or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection between the devices or units may be electrical, mechanical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units; can be located in one place or distributed on a plurality of network units; some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be separately regarded as one unit, or two or more units may be integrated into one unit; the integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: various media capable of storing program codes, such as a removable storage medium, a Read Only Memory (ROM), a magnetic disk, and an optical disk.
Alternatively, the integrated units described above in the present application may be stored in a computer-readable storage medium if they are implemented in the form of software functional modules and sold or used as independent products. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially implemented or portions thereof that contribute to the prior art may be embodied in the form of a software product stored in a storage medium, and including several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a removable storage medium, a ROM, a magnetic disk, an optical disk, or the like, which can store the program code.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. An information processing method based on a block chain is applied to a task initiator for artificial intelligence model training by using the block chain, and the method comprises the following steps of when the issuing condition of the block chain reward is met:
acquiring an incentive value of the block chain incentive and a distribution proportion of the incentive value among a block generator, a first block verifier and a second block verifier, wherein the block generator is a task participant generating a block to update the artificial intelligence model, the first block verifier is a task participant testing the block and contributing local parameters, and the second block verifier is a task participant testing the block but not contributing local parameters;
calculating a first reward value, a second reward value and a third reward value according to the reward value and the distribution proportion;
and issuing the first reward value to the block generator by using an issuing channel provided by the block chain, issuing the second reward value to the first block verifier, and issuing the third reward value to the second block verifier.
2. The method of claim 1, prior to the issuing of the blockchain reward, the method further comprising:
establishing an artificial intelligence model and setting initialization information of the artificial intelligence model, wherein the initialization information comprises an initial model, initial parameters, initial model precision, a precision improvement threshold value and a method for calculating model precision;
determining block generation conditions and block chain reward mechanisms of the block chain, wherein the block chain reward mechanisms comprise distribution conditions of block chain rewards, reward values of the block chain rewards and distribution proportions of the reward values among a block generation party, a first block verification party and a second block verification party;
packing the artificial intelligence model, the initialization information of the artificial intelligence model, the block generation condition and the block chain reward mechanism into a first block of the block chain;
uploading the first block to the block chain.
3. A block chain-based information processing method applied to the block generator of claim 1, the method comprising:
acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as a first model, a first parameter and a first model precision;
updating or not updating the first model as required to obtain a second model, and training the second model by using own data to obtain a second parameter;
according to a pre-obtained method for calculating model precision, calculating second model precision corresponding to the second model and the second parameter and a precision improvement value obtained by comparing the second model precision with the first model precision;
and judging whether the precision improvement value is larger than or equal to a precision improvement threshold value acquired in advance, if so, broadcasting the second model and the second parameter to a block verifier to obtain a voting result of the block verifier, judging whether a block generation condition acquired in advance is met according to the voting result, if so, generating a new block, and acquiring the reward issued to the block generator when the issuing condition of block chain reward is met.
4. The method of claim 3, the block verifier comprising a first block verifier and a second block verifier of claim 1,
accordingly, the broadcasting the second model and the second parameter to the block verifier to obtain the voting result of the block verifier further comprises:
the broadcasting the second model and the second parameter to the block verifier to obtain the local parameter contributed by the first block verifier;
combining the second parameter and a third parameter contributed by the first block verifier to obtain a fourth parameter;
the broadcasting of the second model and the fourth parameter to a block verifier to obtain a voting result of the block verifier.
5. The method of claim 3, prior to obtaining the most recent model, the most recent parameters, and the most recent precision of the artificial intelligence model from the blockchain, the method further comprising:
and acquiring the accuracy of the calculation model, the accuracy improvement threshold and the block generation condition.
6. A block chain-based information processing method applied to the first block verifier of claim 1, the method comprising:
acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as a first model, a first parameter and a first model precision;
receiving a second model and the second parameters broadcast by the tile generator;
training the second model by using own data to obtain a third parameter;
calculating second model precision corresponding to the second model and the third parameter according to a method for calculating model precision acquired in advance;
judging whether the precision of the second model is greater than that of the first model, if so, returning a favorable voting result and returning the third parameter;
and if the block generator can generate a new block, acquiring the reward issued to the first block verifier when the issuing condition of the block chain reward is met.
7. An information processing device based on a block chain, which is applied to a task initiator for artificial intelligence model training by using the block chain, and the device comprises:
the incentive mechanism obtaining module is used for obtaining an incentive value of the block chain incentive and a distribution proportion of the incentive value among a block generating party, a first block verifying party and a second block verifying party, wherein the block generating party is a task participating party for generating a block to update the artificial intelligence model, the first block verifying party is a task participating party for testing the block and contributing local parameters, and the second block verifying party is a task participating party for testing the block but not contributing local parameters;
the reward value distribution module is used for calculating a first reward value, a second reward value and a third reward value according to the reward value and the distribution proportion;
and the reward value issuing module is used for issuing the first reward value to the block generator by using an issuing channel provided by the block chain, issuing the second reward value to the first block verifier and issuing the third reward value to the second block verifier.
8. An information processing apparatus based on a block chain, applied to the block generator of claim 1, the apparatus comprising:
the block acquisition module is used for acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as the first model, the first parameter and the first model precision;
the model training module is used for updating or not updating the first model as required to obtain a second model, and training the second model by using own data to obtain a second parameter;
the precision calculation module is used for calculating second model precision corresponding to the second model and the second parameter and a precision improvement value obtained by comparing the second model precision with the first model precision according to a pre-obtained method for calculating the model precision;
and the block generation module is used for judging whether the precision improvement value is greater than or equal to a pre-acquired precision improvement threshold value, if so, broadcasting the second model and the second parameter to a block verifier to obtain a voting result of the block verifier, judging whether a pre-acquired block generation condition is met according to the voting result, if so, generating a new block, and acquiring the reward issued to the block generator when the issuing condition of the block chain reward is met.
9. An information processing apparatus based on a block chain, applied to the first block verifier of claim 1, the apparatus comprising:
the block acquisition module is used for acquiring the latest model, the latest parameter and the latest precision of the artificial intelligence model from the block chain as the first model, the first parameter and the first model precision;
a receiving module, configured to receive a second model and the second parameter broadcasted by the tile generator;
the model training module is used for training the second model by using own data to obtain a third parameter;
the precision calculation module is used for calculating the precision of a second model corresponding to the second model and the third parameter according to a method for calculating the precision of the model acquired in advance;
the training result feedback module is used for judging whether the precision of the second model is greater than that of the first model, and if so, returning a favorable voting result and returning the third parameter;
and the reward acquisition module is used for acquiring the reward issued to the first block verification party when the block chain reward issuing condition is met if the block generator can generate a new block.
10. A blockchain-based information processing system, the system comprising:
a task initiator for artificial intelligence model training using blockchains for performing the method of claim 1;
a block generator for performing the method of claim 3;
a first block verifier for performing the method of claim 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011061217.5A CN112328610A (en) | 2020-09-30 | 2020-09-30 | Information processing method and device based on block chain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011061217.5A CN112328610A (en) | 2020-09-30 | 2020-09-30 | Information processing method and device based on block chain |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112328610A true CN112328610A (en) | 2021-02-05 |
Family
ID=74313309
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011061217.5A Pending CN112328610A (en) | 2020-09-30 | 2020-09-30 | Information processing method and device based on block chain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112328610A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108491266A (en) * | 2018-03-09 | 2018-09-04 | 联想(北京)有限公司 | Data processing method, device based on block chain and electronic equipment |
| CN109921909A (en) * | 2019-02-15 | 2019-06-21 | 北京工业大学 | The block chain common recognition method and device proved based on contribution |
| US20200193292A1 (en) * | 2018-12-04 | 2020-06-18 | Jinan University | Auditable privacy protection deep learning platform construction method based on block chain incentive mechanism |
| US20200250511A1 (en) * | 2019-02-01 | 2020-08-06 | Zhaoyang Hu | Artist comprehensive ability evaluation and cultivation assistant system based on artificial intelligence |
-
2020
- 2020-09-30 CN CN202011061217.5A patent/CN112328610A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108491266A (en) * | 2018-03-09 | 2018-09-04 | 联想(北京)有限公司 | Data processing method, device based on block chain and electronic equipment |
| US20200193292A1 (en) * | 2018-12-04 | 2020-06-18 | Jinan University | Auditable privacy protection deep learning platform construction method based on block chain incentive mechanism |
| US20200250511A1 (en) * | 2019-02-01 | 2020-08-06 | Zhaoyang Hu | Artist comprehensive ability evaluation and cultivation assistant system based on artificial intelligence |
| CN109921909A (en) * | 2019-02-15 | 2019-06-21 | 北京工业大学 | The block chain common recognition method and device proved based on contribution |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Adler et al. | Astraea: A decentralized blockchain oracle | |
| CN112784994B (en) | Block chain-based federated learning data participant contribution value calculation and excitation method | |
| CN109034892B (en) | Resource allocation method, device, computer readable storage medium and computer equipment | |
| Gordon et al. | Advertising competition in presidential elections | |
| US11714679B2 (en) | Trade platform with reinforcement learning network and matching engine | |
| CN113657608A (en) | Excitation-driven block chain federal learning method | |
| CN109615200B (en) | Virtual interactive prop issuing method, storage medium, electronic device and system | |
| CN108647974A (en) | A kind of Information Authentication method, apparatus and system based on block chain | |
| CN108241970B (en) | Mining method and device based on block chain and computer readable storage medium | |
| CN114372589A (en) | Federated learning method and related device | |
| CN108416619A (en) | A kind of consumption interval time prediction technique, device and readable storage medium storing program for executing | |
| Mathevet et al. | Reputation for a Degree of Honesty | |
| CN114169387A (en) | Model training method and device, electronic equipment and storage medium | |
| CN112328610A (en) | Information processing method and device based on block chain | |
| CN104933593A (en) | Integration excitation system based on Elo hierarchy model and integration excitation method based on Elo hierarchy model | |
| CN110365671B (en) | Crowd sensing incentive mechanism method supporting privacy protection | |
| CN110874609B (en) | User clustering method, storage medium, device and system based on user behaviors | |
| CN115687526A (en) | Seismic data model sharing method based on block chain and federal learning | |
| CN113971582A (en) | Method and system for generating advertisement putting plan, electronic device and storage medium | |
| CN110795132A (en) | Third-party application processing method, system and storage medium | |
| Muñoz et al. | Strategies for exploiting trust models in competitive multi-agent systems | |
| CN110675197A (en) | Method, device and equipment for evaluating data and storage medium | |
| CN112348565A (en) | Information processing method and device based on block chain | |
| Bakhshandeh et al. | Development of reinforcement learning algorithm to study the capacity withholding in electricity energy markets | |
| CN116630550B (en) | Three-dimensional model generation method and system based on multiple pictures |
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 |