阅读说明：本技术 一种基于区块链智能合约的神经网络训练方法及系统 (Neural network training method and system based on block chain intelligent contracts ) 是由 王宏俊 于 2021-09-15 设计创作，主要内容包括：本发明公开了一种基于区块链智能合约的神经网络训练方法及系统,包括包括代币模块、出价模块、信息模块、拟合模块,所述出价模块与信息模块、拟合模块电连接；所述代币模块用于管理代币的充值与退回,所述出价模块用于对竞买人的出价进行收集与运算,所述信息模块用于对竞买人进行信息提供,所述拟合模块用于对竞买人们的出价进行拟合,所述代币模块包括代币充值模块、保证金收取模块、保证金退回模块、代币支付模块,所述出价模块包括出价选择模块、出价统计模块、出价差值记录模块、出价排序模块,所述信息模块包括信息发送模块、信息显示模块、信息有效性计算模块、信息排序模块,本发明,具有保护艺术家隐私、减少不理性抬价的特点。(The invention discloses a neural network training method and a system based on a block chain intelligent contract, which comprises a token module, a bid module, an information module and a fitting module, wherein the bid module is electrically connected with the information module and the fitting module; the system comprises a token module, a bid module, an information module, a fitting module and a bid sorting module, wherein the token module is used for managing recharging and returning of tokens, the bid module is used for collecting and calculating bids of bidders, the information module is used for providing information for the bidders, the fitting module is used for fitting the bids of the bidders, the token module comprises a token recharging module, a guaranteed fund receiving module, a guaranteed fund returning module and a token payment module, the bid module comprises a bid selection module, a bid counting module, a bid difference value recording module and a bid sorting module, and the information module comprises an information sending module, an information display module, an information effectiveness calculating module and an information sorting module.)

1. A neural network training method and system based on a block chain intelligent contract are characterized in that: the system comprises a token module, a bid module, an information module and a fitting module, wherein the bid module is electrically connected with the information module and the fitting module;

the token module is used for managing recharging and returning of the token, the bid module is used for collecting and operating bids of bidders, the information module is used for providing information for the bidders, and the fitting module is used for fitting the bids of the bidders.

2. The neural network training method and system based on the block chain intelligent contract according to claim 1, wherein: the token module comprises a token recharging module, a guarantee fund collecting module, a guarantee fund returning module and a token payment module, wherein the guarantee fund collecting module is electrically connected with the guarantee fund returning module;

the token recharging module is used for recharging tokens for bidders, the deposit collecting module is used for paying the tokens by the bidders to serve as deposit, the deposit returning module is used for returning the deposit after the auction is ended, and the token paying module is used for paying the cost of successful bidders.

3. The neural network training method and system based on the block chain intelligent contract according to claim 2, wherein: the bid module comprises a bid selection module, a bid statistic module, a bid difference value recording module and a bid sorting module, and the bid statistic module is electrically connected with the bid difference value recording module and the bid sorting module;

the bid selection module is used for enabling bidders to select bids, the bid statistics module is used for counting bids of all bidders, the bid difference value recording module is used for recording the difference value of each bid of the bidders, and the bid ordering module is used for ordering the bidders according to the bid heights.

4. The neural network training method and system based on the block chain intelligent contract according to claim 3, wherein: the information module comprises an information sending module, an information display module, an information effectiveness calculation module and an information sorting module, wherein the information sending module is electrically connected with the information display module, and the information effectiveness calculation module is electrically connected with the information sorting module and the bid difference value recording module;

the information display module is used for displaying the related information of the artworks, the information effectiveness calculation module is used for judging information according to the difference value of each bid price of the bidders, and the information sorting module is used for sorting the information according to the effectiveness of the information.

5. The neural network training method and system based on the block chain intelligent contract according to claim 4, wherein: the fitting module comprises a data acquisition module, an operation module and a prediction module, and the operation module is electrically connected with the prediction module;

the data acquisition module is used for acquiring bidding data of bidders, the operation module is used for performing fitting operation on the bidding data, and the prediction module is used for predicting reasonable bidding.

6. The neural network training method and system based on the block chain intelligent contract according to claim 5, wherein: the main work flow of the system is as follows:

s0, before the auction of the electronic artwork starts, the bidder needs to recharge the token through the token recharging module, and the recharged token is used for paying the deposit through the deposit collecting module, the token is used as the deposit and is a guarantee form made by the bidder, so that the bidder is prevented from inadvertently not participating in the appointed auction activity, and the loss of default behaviors to the offeror in all aspects is compensated;

s1, firstly, all bidders have the same mastery degree on the information of the artworks to be auctioned, the bidders bid in the first round according to personal wishes, the bids are given through the bid selection module, the system counts the bids of all bidders through the bid statistics module, and the bidders are ranked through the bid ranking module according to the bid heights, the bidders with high bids can obtain more information related to the artworks, the information is provided for the bidders through the information sending module, and the bidders can check through the information display module;

s2, bidding in the second round, adjusting the price of the bidder according to the related information of the artwork provided by the system after the bidding in the previous round, recording the bidding difference value of the bidder by the system through a bidding difference value recording module, judging what information attracts the bidder according to the bidding difference value of the bidder through an information effectiveness calculating module, sequencing the information according to effectiveness through an information sequencing module, sequencing the bidding in the second round of the bidder through the system, and preferentially providing the information with higher effectiveness to the bidder;

s3, the system collects the data bid twice by the bidders through the data acquisition module, the data is fitted through the operation module, and the reasonable price of the artwork is predicted through the prediction module;

s4, bidding for the third round, wherein the bidder bids for the last time according to the information provided after bidding for the second round and the data predicted by fitting;

s5, the bidder with the highest bid in the third round succeeds in auction, the remaining tokens to be paid are paid out through the token payment module, and the remaining bidders return the deposit through the deposit return module.

7. The neural network training method and system based on the block chain intelligent contract according to claim 6, wherein: in step S1, the rule of information sharing is:

the system can provide more information for bidders with higher bids and provide a small amount of information for bidders with low bids, and the specific method is as follows;

a certain electronic artwork has m pieces of information, n bidders participate in the auction, and the bid price of each bidder in the first round is A_iThe bid price of each bidder in the second round is B_iWhere i ∈ [1, n ]]Then the average bid of the first round bidder is

The relation between the first round bid of the bidder and the information obtained after the first round bid of the bidder is

Where q is the number of messages obtained by the bidder after the first bid round and μ is the adjustment factor, q messages will be randomly drawn from the total of m messages.

8. The neural network training method and system based on the block chain intelligent contract according to claim 7, wherein: in step S2, the specific method for calculating the validity of the information includes:

the initial validity of each piece of information is 100%, and each bidder received q before the second round of bidding_iInformation and according to the information, a new bid B is given_iThen each bidder evaluates the overall validity of the information he receives as

The information is randomly provided for the bidders, the number of times a piece of information is provided for a bidder is variable, the effectiveness of a piece of information is expressed as an average of the total effectiveness evaluations of all bidders receiving the piece of information, and the effectiveness of a piece of information is

Where W is the validity of the information,k is the total of the total effectiveness ratings for all bidders receiving the piece of information and k is the number of bidders receiving the piece of information.

9. The neural network training method and system based on the block chain intelligent contract according to claim 8, wherein: in the step S3, the fitting method specifically includes:

arranging the data of the second round of bidding from low to high, constructing a coordinate system taking the bidding rank as a horizontal axis and the bidding rank as a vertical axis, and fitting the data into a continuous curve through an operation module for reference of bidders;

when the data of the highest bid is higher than the fitted curve, the highest bid is not rational enough, and the price can be reduced properly; when the data of the highest bid overlaps the fitted curve, then this highest bid is more appropriate; when the data for the highest bid is lower than the curve fitted, then this highest bid is lower, and the probability of auction failure in the next round of bidding is higher.

10. The neural network training method and system based on the block chain intelligent contract according to claim 9, wherein: in step S2, the specific method for providing the information with higher effectiveness to the bidder with higher bid includes: the number of information available to the bidder after the second bid isAll information is sorted from high to low in effectiveness, and the information with the highest effectiveness is preferentially provided to the bidders, i.e. each bidder can obtain an effectiveness first high to effectiveness first p_iHigh information.

Technical Field

The invention relates to the technical field of neural networks, in particular to a neural network training method and a system based on a block chain intelligent contract.

Background

The intelligent contract is a protocol defined in a digital form based on a block chain, and provides a safe and reliable transaction mode in internet transactions. Neural networks are machine learning algorithms that have been used for a variety of tasks, including data fitting, computer vision, speech recognition, machine translation, and the like.

The electronic artwork is an electronic edition artwork circulating on the internet and stored in a data form, so that all people can download the artwork, but the ownership of the artwork only belongs to one person, and the ownership of the artwork can be used for auction. During the auction process of electronic artworks, the information related to the artworks is published in the internet very easily, so that the privacy of the artists is damaged. However, the bidders usually want to master more information about the works of art to make more rational decisions, and how to balance the contradiction between these two points becomes a difficult point in the trading of electronic works of art. Therefore, it is necessary to design a neural network training method and system based on a block chain intelligent contract for protecting the information of the artwork.

Disclosure of Invention

The invention aims to provide a neural network training method and system based on a block chain intelligent contract, so as to solve the problems in the background technology.

In order to solve the technical problems, the invention provides the following technical scheme: a neural network training method and system based on a block chain intelligent contract comprises a token module, a bid module, an information module and a fitting module, wherein the bid module is electrically connected with the information module and the fitting module;

the token module is used for managing recharging and returning of the token, the bid module is used for collecting and operating bids of bidders, the information module is used for providing information for the bidders, and the fitting module is used for fitting the bids of the bidders.

According to the technical scheme, the token module comprises a token recharging module, a guarantee fund collecting module, a guarantee fund returning module and a token payment module, wherein the guarantee fund collecting module is electrically connected with the guarantee fund returning module;

the token recharging module is used for recharging tokens for bidders, the deposit collecting module is used for paying the tokens by the bidders to serve as deposit, the deposit returning module is used for returning the deposit after the auction is ended, and the token paying module is used for paying the cost of successful bidders.

According to the technical scheme, the bid module comprises a bid selection module, a bid statistic module, a bid difference value recording module and a bid sorting module, wherein the bid statistic module is electrically connected with the bid difference value recording module and the bid sorting module;

the bid selection module is used for enabling bidders to select bids, the bid statistics module is used for counting bids of all bidders, the bid difference value recording module is used for recording the difference value of each bid of the bidders, and the bid ordering module is used for ordering the bidders according to the bid heights.

According to the technical scheme, the information module comprises an information sending module, an information display module, an information effectiveness calculating module and an information sorting module, wherein the information sending module is electrically connected with the information display module, and the information effectiveness calculating module is electrically connected with the information sorting module and the bid difference value recording module;

the information display module is used for displaying the related information of the artworks, the information effectiveness calculation module is used for judging information according to the difference value of each bid price of the bidders, and the information sorting module is used for sorting the information according to the effectiveness of the information.

According to the technical scheme, the fitting module comprises a data acquisition module, an operation module and a prediction module, wherein the operation module is electrically connected with the prediction module;

the data acquisition module is used for acquiring bidding data of bidders, the operation module is used for performing fitting operation on the bidding data, and the prediction module is used for predicting reasonable bidding.

According to the technical scheme, the main work flow of the system is as follows:

s0, before the auction of the electronic artwork starts, the bidder needs to recharge the token through the token recharging module, and the recharged token is used for paying the deposit through the deposit collecting module, the token is used as the deposit and is a guarantee form made by the bidder, so that the bidder is prevented from inadvertently not participating in the appointed auction activity, and the loss of default behaviors to the offeror in all aspects is compensated;

s1, firstly, all bidders have the same mastery degree on the information of the artworks to be auctioned, the bidders bid in the first round according to personal wishes, the bids are given through the bid selection module, the system counts the bids of all bidders through the bid statistics module, and the bidders are ranked through the bid ranking module according to the bid heights, the bidders with high bids can obtain more information related to the artworks, the information is provided for the bidders through the information sending module, and the bidders can check through the information display module;

s2, bidding in the second round, adjusting the price of the bidder according to the related information of the artwork provided by the system after the bidding in the previous round, recording the bidding difference value of the bidder by the system through a bidding difference value recording module, judging what information attracts the bidder according to the bidding difference value of the bidder through an information effectiveness calculating module, sequencing the information according to effectiveness through an information sequencing module, sequencing the bidding in the second round of the bidder through the system, and preferentially providing the information with higher effectiveness to the bidder;

s3, the system collects the data bid twice by the bidders through the data acquisition module, the data is fitted through the operation module, and the reasonable price of the artwork is predicted through the prediction module;

s4, bidding for the third round, wherein the bidder bids for the last time according to the information provided after bidding for the second round and the data predicted by fitting;

s5, the bidder with the highest bid in the third round succeeds in auction, the remaining tokens to be paid are paid out through the token payment module, and the remaining bidders return the deposit through the deposit return module.

According to the above technical solution, in the step S1, the rule of information sharing is:

the system can provide more information for bidders with higher bids and provide a small amount of information for bidders with low bids, and the specific method is as follows;

a certain electronic artwork has m pieces of information, n bidders participate in the auction, and the bid price of each bidder in the first round is A_iThe bid price of each bidder in the second round is B_iWhere i ∈ [1, n ]]Then the average bid of the first round bidder is

The relation between the first round bid of the bidder and the information obtained after the first round bid of the bidder is

Where q is the number of messages obtained by the bidder after the first bid round and μ is the adjustment factor, q messages will be randomly drawn from the total of m messages.

According to the above technical solution, in the step S2, the specific method for calculating the validity of the information is as follows:

the initial validity of each piece of information is 100%, and each bidder received q before the second round of bidding_iInformation and according to the information, a new bid B is given_iThen each bidder evaluates the overall validity of the information he receives as

The information is randomly provided for the bidders, the number of times a piece of information is provided for a bidder is variable, the effectiveness of a piece of information is expressed as an average of the total effectiveness evaluations of all bidders receiving the piece of information, and the effectiveness of a piece of information is

Where W is the validity of the information,k is the total of the total effectiveness ratings for all bidders receiving the piece of information and k is the number of bidders receiving the piece of information.

According to the above technical solution, in the step S3, the fitting method specifically includes:

arranging the data of the second round of bidding from low to high, constructing a coordinate system taking the bidding rank as a horizontal axis and the bidding rank as a vertical axis, and fitting the data into a continuous curve through an operation module for reference of bidders;

when the data of the highest bid is higher than the fitted curve, the highest bid is not rational enough, and the price can be reduced properly; when the data of the highest bid overlaps the fitted curve, then this highest bid is more appropriate; when the data for the highest bid is lower than the curve fitted, then this highest bid is lower, and the probability of auction failure in the next round of bidding is higher.

According to the above technical solution, in the step S2, a specific method for providing the information with higher effectiveness to the bidder having the higher bid includes: the number of information available to the bidder after the second bid isAll information is sorted from high to low in effectiveness, and the information with the highest effectiveness is preferentially provided to the bidders, i.e. each bidder can obtain an effectiveness first high to effectiveness first p_iHigh information.

Compared with the prior art, the invention has the following beneficial effects: according to the invention, the sending quantity of the information is related to the bid of the bidder, so that the information is effectively prevented from being leaked to a person with low willingness on the art auction. Among all information, the information with higher effectiveness can be seen by most people, and the information with lower effectiveness can hardly be seen by people, so that the information irrelevant to the artwork auction is protected from being acquired by most people. In the auction, the bidders with higher bids can see more comprehensive information, so that the bidders can make more rational decisions, and the bidders are prevented from misjudging the value of the artwork due to one side of information mastering. By arranging the fitting module, the bidding data of the bidders are fitted into a curve, so that the bidders can more intuitively evaluate the bids of the bidders, and the bidders can be helped to give more reasonable prices in the last bidding round.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:

FIG. 1 is a schematic view of the overall module structure of the present invention;

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all 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 invention.

As shown in fig. 1, the present invention provides the technical solution: a neural network training method and system based on a block chain intelligent contract comprises a token module, a bid module, an information module and a fitting module, wherein the bid module is electrically connected with the information module and the fitting module;

the system comprises a token module, a bid module, an information module and a fitting module, wherein the token module is used for managing recharging and returning of tokens, the bid module is used for collecting and calculating bids of bidders, the information module is used for providing information to the bidders, and the fitting module is used for fitting the bids of the bidders;

the token module comprises a token recharging module, a guarantee fund collecting module, a guarantee fund returning module and a token payment module, wherein the guarantee fund collecting module is electrically connected with the guarantee fund returning module;

the system comprises a token recharging module, a deposit fund collection module, a deposit fund withdrawal module and a token payment module, wherein the token recharging module is used for recharging tokens for bidders, the deposit fund collection module is used for paying the tokens by the bidders as deposit funds, the deposit fund withdrawal module is used for withdrawing the deposit funds after the auction is ended, and the token payment module is used for paying the fees of successful bidders;

the bid module comprises a bid selection module, a bid statistic module, a bid difference value recording module and a bid sorting module, and the bid statistic module is electrically connected with the bid difference value recording module and the bid sorting module;

the bid selection module is used for enabling bidders to select bids, the bid statistics module is used for counting the bids of all bidders, the bid difference value recording module is used for recording the difference value of each bid of the bidders, and the bid ordering module is used for ordering the bidders according to the bid heights;

the information module comprises an information sending module, an information display module, an information effectiveness calculation module and an information sorting module, wherein the information sending module is electrically connected with the information display module, and the information effectiveness calculation module is electrically connected with the information sorting module and the bid difference value recording module;

the information display module is used for displaying the related information of the artwork, the information effectiveness calculation module is used for judging information according to the difference value of each bid price of the bidders, and the information sorting module is used for sorting the information according to the effectiveness of the information;

the fitting module comprises a data acquisition module, an operation module and a prediction module, and the operation module is electrically connected with the prediction module;

the data acquisition module is used for acquiring bidding data of bidders, the operation module is used for performing fitting operation on the bidding data, and the prediction module is used for predicting reasonable bidding;

the main work flow of the system is as follows:

s0, before the auction of the electronic artwork starts, the bidder needs to recharge the token through the token recharging module, and the recharged token is used for paying the deposit through the deposit collecting module, the token is used as the deposit and is a guarantee form made by the bidder, so that the bidder is prevented from inadvertently not participating in the appointed auction activity, and the loss of default behaviors to the offeror in all aspects is compensated;

s1, firstly, all bidders have the same mastery degree on the information of the artworks to be auctioned, the bidders bid in the first round according to personal wishes, the bids are given through the bid selection module, the system counts the bids of all bidders through the bid statistics module, and the bidders are ranked through the bid ranking module according to the bid heights, the bidders with high bids can obtain more information related to the artworks, the information is provided for the bidders through the information sending module, and the bidders can check through the information display module;

s2, bidding in the second round, adjusting the price of the bidder according to the related information of the artwork provided by the system after the bidding in the previous round, recording the bidding difference value of the bidder by the system through a bidding difference value recording module, judging what information attracts the bidder according to the bidding difference value of the bidder through an information effectiveness calculating module, sequencing the information according to effectiveness through an information sequencing module, sequencing the bidding in the second round of the bidder through the system, and preferentially providing the information with higher effectiveness to the bidder;

s3, the system collects the data bid twice by the bidders through the data acquisition module, the data is fitted through the operation module, and the reasonable price of the artwork is predicted through the prediction module;

s4, bidding for the third round, wherein the bidder bids for the last time according to the information provided after bidding for the second round and the data predicted by fitting;

s5, the bidder with the highest bid in the third round successfully takes the auction, the remaining tokens to be paid are paid by the token payment module, and the remaining bidders return the deposit by the deposit return module;

through the steps, information is effectively prevented from being leaked to a person with low intention on the art auction, and more rational decision making is facilitated for the bidders;

in step S1, the rule of information sharing is:

the system can provide more information for bidders with higher bids and provide a small amount of information for bidders with low bids, and the specific method is as follows;

a certain electronic artwork has m pieces of information, n bidders participate in the auction, and the bid price of each bidder in the first round is A_iThe bid price of each bidder in the second round is B_iWhere i ∈ [1, n ]]Then the average bid of the first round bidder is

The relation between the first round bid of the bidder and the information obtained after the first round bid of the bidder is

Wherein q is the number of information obtained by the bidder after the first round of bidding, mu is an adjustment coefficient, and q pieces of information are randomly extracted from the total m pieces of information;

in step S2, the specific method for calculating the validity of the information includes:

the initial validity of each piece of information is 100%, and each bidder received q before the second round of bidding_iInformation and according to the information, a new bid B is given_iThen each bidder evaluates the overall validity of the information he receives as

The information is randomly provided for the bidders, the number of times a piece of information is provided for a bidder is variable, the effectiveness of a piece of information is expressed as an average of the total effectiveness evaluations of all bidders receiving the piece of information, and the effectiveness of a piece of information is

Where W is the validity of the information,k is the total sum of the total effectiveness evaluation of all the bidders receiving the information, and k is the number of the bidders receiving the information;

in the step S3, the fitting method specifically includes:

arranging the data of the second round of bidding from low to high, constructing a coordinate system taking the bidding rank as a horizontal axis and the bidding rank as a vertical axis, and fitting the data into a continuous curve through an operation module for reference of bidders;

when the data of the highest bid is higher than the fitted curve, the highest bid is not rational enough, and the price can be reduced properly; when the data of the highest bid overlaps the fitted curve, then this highest bid is more appropriate; when the data for the highest bid is lower than the curve fitted, then this highest bid is lower, the probability of auction failure in the next round of bidding is higher;

by arranging the fitting module, the bidding data of the bidders are fitted into a curve, so that the bidders can more intuitively evaluate the bids of the bidders, and the bidders can be helped to give more reasonable prices in the last bidding round;

in step S2, the specific method for providing the information with higher effectiveness to the bidder with higher bid includes: the number of information available to the bidder after the second bid isAll information is sorted from high to low in effectiveness, and the information with the highest effectiveness is preferentially provided to the bidders, i.e. each bidder can obtain an effectiveness first high to effectiveness first p_iHigh information.

Example (b):

the bidder a first needs to recharge the token and pay the deposit before the auction. The artwork has 20 pieces of information, and 10 bidders participate in the auction. In the first round of bidding, bidder a gave a bid of 10000, the average bid for all bidders was 6000, and the bid for bidder a was the highest. Assuming that the adjustment coefficient mu is 3, according to the formulaIt is available that after the first round of bidding bidder a can randomly obtain 10 pieces of information.

The second round of bidding, bidder a has made a price adjustment, giving a bid of 10500, the average bid of all bidders is 7000, and the bid of bidder a is highest. According to the formulaAfter the bidder a has bid, the validity of the 10 pieces of information he has obtained becomes 120%. Wherein a certain piece of information is provided to 4 bidders according to a formulaThe effectiveness of this information is an average of the effectiveness scores given by the four bidders. All information is ranked from high to low in effectiveness, and bidder a can obtain 9 pieces of information from the first to the 9 th in effectiveness after the second round of bidding.

The system, by fitting the bid data, the bidder a finds that his bid 10500 is above the curve and the bid is too high. Thus in the third round of bidding, bidder a has given a bid 9000 based on the information provided and the fitted curve for reference, and the auction is successful.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, 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.

Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.