Multi-airport data collaborative prediction method and system based on privacy calculation

Abstract

The invention belongs to the technical field of airport data prediction, and discloses a multi-airport data collaborative prediction method and system based on privacy calculation. According to the method, model parameters are set at a central server side, a local data set is arranged according to a general prediction model, data statistics is carried out, and after data statistics is completed, a multiple linear regression model is used for training; the central server side is responsible for planning training of all participants as a whole and generating a final aggregation model; and the server side uses a federated average algorithm to carry out security aggregation on the regression coefficients and then issues the updated parameter regression coefficients to each client side, and the client sides adopt the updated parameters during actual prediction. According to the invention, privacy protection in the data sharing process is realized, and the risk of data leakage is reduced. And the original data of the airport is not out of the domain and is available and invisible.

Description

Technical Field

[0001] The present invention belongs to the technical field of airport data prediction, and in particular relates to a multi-airport data collaborative prediction method and system based on privacy computing. Background Art

[0002] With the rapid development of airport informatization, airports, as modern transportation hubs, have assumed increasingly important traffic and transportation tasks, and have also become important information centers and data processing centers. Currently, airport data is growing explosively, with huge amounts of data and diversified data sources, including flight dynamics, passenger travel, air logistics, etc., which are of great value to social development. However, when using data from one airport for predictive analysis, there are often problems such as small data sample size that is insufficient to support model training and low prediction accuracy. When using data from multiple airports for predictive analysis, each airport needs to share data, which can easily lead to privacy leaks and data abuse, and the data cannot be circulated and used safely and reliably.

[0003] Through the above analysis, the problems and defects of the existing technology are as follows: when dealing with the problem of airport collaborative prediction and analysis, current technology often adopts methods such as plaintext data sharing and data encryption and sharing to achieve it. However, the risk of plaintext data sharing is very high and data privacy protection cannot be achieved; although data can be shared after encryption, the data demander still needs to decrypt it into plaintext before using it, and there is still a risk of data privacy leakage. Therefore, how to achieve predictive analysis of multi-airport data in an invisible way has become an urgent problem that needs to be solved. Summary of the invention

[0004] In order to overcome the problems existing in the related technologies, the disclosed embodiments of the present invention provide a multi-airport data collaborative prediction method and system based on privacy computing. The present invention adopts advanced privacy protection algorithms and technical means to ensure the circulation and use of multi-airport data in a safe environment, protect personal privacy and data security, and solve the problem of data security and compliance use.

[0005] The technical solution is as follows: a multi-airport data collaborative prediction method based on privacy computing, comprising: S1, set the parameters of the airport data general prediction model based on the multivariate linear regression algorithm on the central server, including the dependent variable and independent variables and distribute them to the client nodes participating in the training at each airport; S2, after receiving the parameters of the universal prediction model, each airport client node organizes the local data set according to the universal prediction model and performs data statistics. After the data statistics are completed, the multivariate linear regression model is used for training; S3, the central server aggregates and returns to the client. After the local training of each airport client is completed, each regression coefficient is uploaded to the central server. The central server coordinates the training of all participants and generates the final aggregate model. S4, local airport algorithm model update, the server uses the federated average algorithm to securely aggregate the regression coefficients, and then sends the updated parameter regression coefficients to each client. The client uses the updated parameters for actual prediction.

[0006] In step S1, the general prediction model adopts the airport store turnover prediction model. The parameters of the airport store turnover prediction model include dependent variables and 12 independent variables, all of which are based on the day as the statistical dimension. The dependent variable is the turnover of all stores in the airport, and the independent variables are domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, domestic inbound passengers, domestic outbound passengers, international inbound passengers, international outbound passengers, the number of flights canceled on the day, the number of flights delayed on the day, date labels, and weather labels. The airports participating in the training are airports.

[0007] In step S2, after receiving the parameters of the general prediction model, each airport client node organizes the local data set according to the general prediction model and performs data statistics, including: After each airport client node receives the parameters of the airport store turnover prediction model, it organizes the local data set according to the parameters of the airport store turnover prediction model, and performs statistical calculations based on the data of the production operation system, passenger service system, security information system, meteorological service system, and airport business system in the airport to generate data on the number of domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, the number of domestic inbound passengers, the number of domestic outbound passengers, the number of international inbound passengers, the number of international outbound passengers, the number of canceled flights on the day, and the number of delayed flights on the day, and generates date labels and weather labels based on the date and weather of the day.

[0008] In step S2, the expression of the multiple linear regression model is: In the formula, is a constant term; for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of .

[0009] In step S3, the central server coordinates the training of all participants and generates the final aggregate model, including: The central server determines the global communication round , complete the system initialization configuration; the central server receives Client aggregation model parameters Using the federated average algorithm for secure aggregation generation , the aggregation model calculation formula is as follows: In the formula, For the The proportion of clients in the entire aggregation model training process is calculated as follows: In the formula, For the The number of data items that each client participates in the training of the aggregation model.

[0010] In step S4, the client uses the updated parameters for actual prediction, including: The updated parameters Issued to An airport client, when the airport makes a store turnover forecast, it uses the parameters aggregated and updated by the central server to make a forecast, and counts the number of domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, the number of domestic inbound passengers, the number of domestic outbound passengers, the number of international inbound passengers, the number of international outbound passengers, the number of canceled flights, the number of delayed flights, date labels, and weather labels for the day, and brings them into the model to predict the store turnover for the day.

[0011] Another object of the present invention is to provide a multi-airport data collaborative prediction system based on privacy computing, which implements the multi-airport data collaborative prediction method based on privacy computing. The system includes: The model parameter setting module is used to set the parameters of the airport data general prediction model based on the multivariate linear regression algorithm on the central server, including the dependent variable and independent variables and distribute them to the client nodes participating in the training at each airport; The multivariate linear regression training module is used for each airport client node to receive the parameters of the general prediction model, organize the local data set according to the general prediction model, perform data statistics, and use the multivariate linear regression model for training after the data statistics are completed; The aggregation model generation module is used for the central server to aggregate and return to the client. After the local training of each airport client is completed, the respective regression coefficients are uploaded to the central server. The central server coordinates the training of all participants and generates the final aggregation model. The parameter prediction module is used to update the local airport algorithm model. After the server uses the federated average algorithm to securely aggregate the regression coefficients, the updated parameter regression coefficients are sent to each client. The client uses the updated parameters for actual prediction.

[0012] Furthermore, the system is mounted on a computer device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, and when the processor executes the computer program, the functions of the multi-airport data collaborative prediction system based on privacy computing are implemented.

[0013] Furthermore, the system is equipped with a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can realize the functions of the multi-airport data collaborative prediction system based on privacy computing.

[0014] Furthermore, the system is equipped with an information data processing terminal, which is used to provide a user input interface when executed on an electronic device to implement functions such as the multi-airport data collaborative prediction system based on privacy computing mentioned above. The information data processing terminal includes a mobile phone, a computer, and a switch.

[0015] Combining all the above technical solutions, the beneficial effects of the present invention are as follows: Based on the privacy computing technology of federated learning, the present invention enables each airport to complete data prediction and analysis without exposing their own data, realizes privacy protection in the process of data sharing and reduces the risk of data leakage. The original data of the airport does not leave the domain, the data is available but not visible, and new technologies and methods are used to help the safe and orderly circulation and use of airport data, thereby promoting the data factorization level of the airport to a new height, and better serving the operation of the airport and the travel of passengers.

[0016] The present invention combines data from multiple airports for integrated analysis, making data samples richer and more comprehensive, helping to discover potential patterns and trends, improving the accuracy and reliability of prediction models, and more accurately guiding and improving airport operations. Multiple airports can jointly conduct data analysis and prediction without leaking their own sensitive data, greatly improving data security, reducing the risk of data leakage, and protecting the privacy rights and interests of airports, passengers, and related stakeholders.

[0017] The data used in existing prediction methods often come from a single airport. Due to the small data sample, the prediction results may be inaccurate, unrepresentative, and have sampling bias. This method can use multi-airport data for integrated analysis. The large data sample is more helpful in discovering potential patterns and trends and improving the accuracy and reliability of the prediction model. This method uses privacy computing technology based on federated learning to ensure that in the process of data collaborative prediction, the original data does not go beyond the control of the original data owner, achieving "data available but invisible". This means that multiple participating airports can jointly conduct data analysis and prediction without leaking their own sensitive data, which greatly improves data security, reduces the risk of data leakage, and protects the privacy of airports, passengers, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure; Figure 1 is a flow chart of a multi-airport data collaborative prediction method based on privacy computing provided by an embodiment of the present invention; Figure 2 is a schematic diagram of a multi-airport data collaborative prediction system based on privacy computing provided by an embodiment of the present invention; In the figure: 1. Model parameter setting module; 2. Multiple linear regression training module; 3. Aggregation model generation module; 4. Parameter prediction module. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0020] The innovation of the present invention lies in: the present invention is based on the privacy computing technology of federated learning, which has the characteristics of "data remains unchanged, model moves", so that each airport can complete data collaborative prediction and analysis without exposing their own data. The data comes from multiple airports, making the data samples richer and more comprehensive, improving the accuracy and reliability of the prediction model, and the airports do not need to share their own sensitive data, which greatly improves data security, reduces the risk of data leakage, and protects the privacy rights and interests of airports, passengers, etc.

[0021] Example 1: Data factorization is the key to releasing the value of data. By studying privacy computing technology, the original data of the airport can be kept within the domain, the data can be available but not visible, and data privacy can be effectively protected. Privacy computing is a cutting-edge interdisciplinary subject that aims to protect data privacy while realizing the transformation and circulation of data value. With the advent of the big data era, data has become a new factor of production, and the need for data privacy protection has become increasingly prominent. Privacy computing achieves a balance between data availability and privacy protection through a series of technical means, providing a solution for the safe circulation of airport data.

[0022] The multi-airport data collaborative prediction method based on privacy computing provided by the embodiment of the present invention adopts a federated learning algorithm to implement privacy computing based on airport data. Federated learning is an emerging machine learning technology in recent years. It is a distributed machine learning architecture for joint training of multi-party computing nodes, aiming to integrate the knowledge contained in multi-party data, improve system performance, and protect data privacy. The present invention adopts the federated learning architecture as a client-server architecture. In this system, multiple airports with the same data structure first complete the training of the algorithm model locally with their own data, perform model aggregation on the central server, and collaboratively train a machine learning model.

[0023] like Figure 1 As shown, the multi-airport data collaborative prediction method based on privacy computing provided by an embodiment of the present invention includes: S1, set the parameters of the airport data general prediction model based on the multivariate linear regression algorithm on the central server, including the dependent variable and independent variables and distribute them to the client nodes participating in the training at each airport; The general prediction model is used for airport passenger flow prediction, airport store sales prediction, etc. The general prediction model parameters are set on the central server, including dependent variables and independent variables and distribute them to the client nodes participating in the training at each airport.

[0024] S2, after receiving the parameters of the universal prediction model, each airport client node organizes the local data set according to the universal prediction model and performs data statistics. After the data statistics are completed, the multivariate linear regression model is used for training; According to the general prediction, the local data set is sorted and data statistics are performed. After the data statistics are completed, the multivariate linear regression model is used for training. The multivariate linear regression model of each client is expressed as: In the formula, is a constant term; for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of .

[0025] S3, the central server aggregates and returns to the client. After the local training of each airport client is completed, each regression coefficient is uploaded to the central server. The central server coordinates the training of all participants and generates the final aggregate model. The airport client downloads the aggregation parameters again to update the local model and conduct the next iteration of training until the global model converges. The central server determines the global communication round , complete the system initialization configuration; the central server receives Client aggregation model parameters Using the federated average algorithm for secure aggregation generation , the aggregation model calculation formula is as follows: In the formula, For the The proportion of clients in the entire aggregation model training process is calculated as follows: In the formula, For the The number of data items that each client participates in the training of the aggregation model.

[0026] In this algorithm, the contribution of each participant's aggregation model parameters to the aggregation result is calculated based on its own data volume. This method can ensure that participants with larger data volumes have more influence on the update of aggregation model parameters, thereby improving the accuracy and generalization ability of the aggregation model.

[0027] S4, local airport algorithm model update, the server uses the federated average algorithm to securely aggregate the regression coefficients, and then sends the updated parameter regression coefficients to each client. The client uses the updated parameters for actual prediction.

[0028] At the same time, it can be seen from the above embodiments that the present invention is a universal prediction model based on multi-airport data, which has the characteristics of "data does not move, model moves". The difference from the traditional centralized prediction model is that there is no need to aggregate the data of each airport together, and unified prediction model training can be completed, which effectively ensures the privacy of private data of all parties and reduces the risk of privacy leakage. In the federated learning mode, there is no need to share or exchange private data samples between the participants, only the relevant parameters in the model training process need to be exchanged. Federated learning solves the problems in current big data research that the data sample size of a single data holder is too small to support model training, the data holders cannot directly exchange data, and are unwilling to share original data samples.

[0029] Example 2, as another implementation of the present invention, taking the airport store sales prediction model as an example, the multi-airport data collaborative prediction method based on privacy computing provided by the embodiment of the present invention includes: (1) Setting model parameters. The airport data prediction model includes the airport passenger flow prediction model and the airport store turnover prediction model. The prediction algorithm used by the platform is the multiple linear regression algorithm. Here, the airport store turnover prediction model is taken as an example. This model is a prediction model based on the multiple linear regression algorithm. By comprehensively analyzing factors such as the number of passengers, number of flights, and weather at the airport, it completes the prediction of the overall turnover of airport stores, which helps to guide and improve the operation of the airport. Here, the parameters of the airport store turnover prediction model need to be set on the central server, including the dependent variable and 12 independent variables, all of which are based on the day as the statistical dimension. The dependent variable is the turnover of all stores at the airport, and the independent variables are the number of domestic inbound flights, the number of domestic outbound flights, the number of international inbound flights, the number of international outbound flights, the number of domestic inbound passengers, the number of domestic outbound passengers, the number of international inbound passengers, the number of international outbound passengers, the number of flights canceled on the day, the number of flights delayed on the day, date labels (including weekdays, weekends other than holidays, holidays, etc.), and weather labels (including sunny, cloudy, overcast, light rain, moderate rain, heavy rain, light snow, moderate snow, heavy snow, etc.). Here, the airports participating in the training are set as k airports such as Airport A, Airport B, and Airport C, and the model parameters are distributed to the client nodes participating in the training of these k airports.

[0030] (2) Airport local algorithm model training. After receiving the parameters of the airport store turnover prediction model, i.e., the dependent variable and 12 independent variables, the k airport client nodes organize the local data set according to the parameters of the airport store turnover prediction model. According to the historical flight details in the airport production and operation system, the daily domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, the number of flights canceled on the day, the number of flights delayed on the day and other independent variable data can be calculated. According to the historical passenger details in the airport passenger service system and security information system, the daily number of domestic inbound passengers, domestic outbound passengers, international inbound passengers, international outbound passengers and other independent variable data can be calculated. According to the date, the daily date label (including working days, weekends excluding holidays, holidays, etc.) can be generated. According to the historical weather data in the airport meteorological service system, the weather label (including sunny, cloudy, overcast, light rain, moderate rain, heavy rain, light snow, moderate snow, heavy snow, etc.) can be generated. According to the historical store operation data in the airport commercial system, the daily overall store turnover, i.e., the dependent variable data, can be calculated.

[0031] After the data statistics are completed, the multivariate linear regression model is used for training. The model of each client can be expressed as: In the formula, is a constant term; for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of .

[0032] (3) The central server aggregates and returns to the client. After the local training of each airport client is completed, the model parameters of each The central server is responsible for coordinating the training of all participants and aggregating the model parameters of each client to generate , Each airport client downloads the aggregated model parameters and updates the original model parameters to the aggregated model parameters, and then performs the next iteration of training until the global model converges.

[0033] First, the central server determines the global communication round , complete the system initialization configuration. Secondly, the central server needs to receive Client aggregation model parameters Using the federated average algorithm for secure aggregation generation , and its calculation formula is as follows: in, is the proportion of the kth client in the entire aggregation model training process, and its calculation formula is as follows: in, The number of data items participating in the aggregation model training of the k-th client.

[0034] In the federated average algorithm, the contribution of each participant's aggregation model parameters to the aggregation result is calculated based on its own data volume. This method can ensure that participants with larger data volumes have more influence on the update of aggregation model parameters, thereby improving the accuracy and generalization ability of the aggregation model.

[0035] (4) Local airport algorithm model update: the central server uses the federated average algorithm to securely aggregate the regression coefficients and then adds the model parameters aggregated in the previous step to Issued to When the airport predicts store sales, it can use the parameters aggregated and updated by the central server to make predictions, count the number of domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, domestic inbound passengers, domestic outbound passengers, international inbound passengers, international outbound passengers, the number of flights canceled on the day, the number of flights delayed on the day, date labels, and weather labels, and bring them into the multivariate linear regression model based on the aggregated model parameters to predict the store sales on the day.

[0036] Embodiment 3, as Figure 2 As shown, an embodiment of the present invention provides a multi-airport data collaborative prediction system based on privacy computing, the system comprising: Model parameter setting module 1 is used to set the parameters of the airport data general prediction model based on the multivariate linear regression algorithm on the central server side, including dependent variables and independent variables and distribute them to the client nodes participating in the training at each airport; Multiple linear regression training module 2 is used for each airport client node to receive the parameters of the general prediction model, sort out the local data set according to the general prediction model, perform data statistics, and use the multiple linear regression model for training after the data statistics are completed; Aggregation model generation module 3 is used for central server-side aggregation and returns to the client. After the local training of each airport client is completed, the respective regression coefficients are uploaded to the central server. The central server coordinates the training of each participant and generates the final aggregation model. Parameter prediction module 4 is used to update the local airport algorithm model. After the server uses the federated average algorithm to securely aggregate the regression coefficients, the updated parameter regression coefficients are sent to each client. The client uses the updated parameters for actual prediction.

[0037] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.

Claims

1. A multi-airport data collaborative prediction method based on privacy computing, characterized in that: The method includes: S1, set the parameters of the airport data general prediction model based on the multivariate linear regression algorithm on the central server, including the dependent variable and independent variables and distribute them to the client nodes participating in the training at each airport; S2, after receiving the parameters of the universal prediction model, each airport client node organizes the local data set according to the universal prediction model and performs data statistics. After the data statistics are completed, the multivariate linear regression model is used for training; S3, the central server aggregates and returns to the client. After the local training of each airport client is completed, each regression coefficient is uploaded to the central server. The central server coordinates the training of all participants and generates the final aggregate model. S4, local airport algorithm model update, the server uses the federated average algorithm to securely aggregate the regression coefficients, and then sends the updated parameter regression coefficients to each client. The client uses the updated parameters for actual prediction.

2. The multi-airport data collaborative prediction method based on privacy computing according to claim 1 is characterized in that: In step S1, the general prediction model adopts the airport store turnover prediction model. The parameters of the airport store turnover prediction model include dependent variables and 12 independent variables, all of which are based on the day as the statistical dimension. The dependent variable is the turnover of all stores in the airport, and the independent variables are domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, domestic inbound passengers, domestic outbound passengers, international inbound passengers, international outbound passengers, the number of flights canceled on the day, the number of flights delayed on the day, date labels, and weather labels. The airports participating in the training are airports.

3. The multi-airport data collaborative prediction method based on privacy computing according to claim 1 is characterized in that: In step S2, after receiving the parameters of the general prediction model, each airport client node organizes the local data set according to the general prediction model and performs data statistics, including: After each airport client node receives the parameters of the airport store turnover prediction model, it organizes the local data set according to the parameters of the airport store turnover prediction model, and performs statistical calculations based on the data of the production operation system, passenger service system, security information system, meteorological service system, and airport business system in the airport to generate data on the number of domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, the number of domestic inbound passengers, the number of domestic outbound passengers, the number of international inbound passengers, the number of international outbound passengers, the number of canceled flights on the day, and the number of delayed flights on the day, and generates date labels and weather labels based on the date and weather of the day.

4. The multi-airport data collaborative prediction method based on privacy computing according to claim 1 is characterized in that: In step S2, the expression of the multiple linear regression model is: In the formula, is a constant term; for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of for When fixed, Each additional unit The effect of right The partial regression coefficient of .

5. The multi-airport data collaborative prediction method based on privacy computing according to claim 1 is characterized in that: In step S3, the central server coordinates the training of all participants and generates the final aggregate model, including: The central server determines the global communication round , complete the system initialization configuration; the central server receives Client aggregation model parameters Using the federated average algorithm for secure aggregation generation , the aggregation model calculation formula is as follows: In the formula, For the The proportion of clients in the entire aggregation model training process is calculated as follows: In the formula, For the The number of data items that each client participates in the training of the aggregation model.

6. The multi-airport data collaborative prediction method based on privacy computing according to claim 4 is characterized in that: In step S4, the client uses the updated parameters for actual prediction, including: The updated parameters Issued to An airport client, when the airport makes a store turnover forecast, it uses the parameters aggregated and updated by the central server to make a forecast, and counts the number of domestic inbound flights, domestic outbound flights, international inbound flights, international outbound flights, the number of domestic inbound passengers, the number of domestic outbound passengers, the number of international inbound passengers, the number of international outbound passengers, the number of canceled flights, the number of delayed flights, date labels, and weather labels for the day, and brings them into the model to predict the store turnover for the day.

7. A multi-airport data collaborative prediction system based on privacy computing, characterized in that: The system implements the multi-airport data collaborative prediction method based on privacy computing as described in any one of claims 1 to 6, and the system includes: The model parameter setting module (1) is used to set the parameters of the airport data universal prediction model based on the multivariate linear regression algorithm on the central server side, including the dependent variable and independent variables and distribute them to the client nodes participating in the training at each airport; The multivariate linear regression training module (2) is used for each airport client node to receive the parameters of the general prediction model, sort out the local data set according to the general prediction model, perform data statistics, and use the multivariate linear regression model for training after the data statistics are completed; The aggregation model generation module (3) is used for the central server to aggregate and return to the client. After the local training of each airport client is completed, the respective regression coefficients are uploaded to the central server. The central server coordinates the training of each participant and generates the final aggregation model. The parameter prediction module (4) is used to update the local airport algorithm model. After the server uses the federated average algorithm to securely aggregate the regression coefficients, the updated parameter regression coefficients are sent to each client. The client uses the updated parameters when making actual predictions.

8. The multi-airport data collaborative prediction system based on privacy computing according to claim 7 is characterized in that: The system is mounted on a computer device, which includes: at least one processor, a memory, and a computer program stored in the memory and running on the at least one processor. When the processor executes the computer program, the functions of the multi-airport data collaborative prediction system based on privacy computing are implemented.

9. The multi-airport data collaborative prediction system based on privacy computing according to claim 7 is characterized in that: The system is equipped with a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can realize the functions of the multi-airport data collaborative prediction system based on privacy computing.

10. The multi-airport data collaborative prediction system based on privacy computing according to claim 7 is characterized in that: The system is equipped with an information data processing terminal, which is used to provide a user input interface when executed on an electronic device to implement functions such as the multi-airport data collaborative prediction system based on privacy computing mentioned above. The information data processing terminal includes a mobile phone, a computer, and a switch.

Images