AI big data analysis-based exchange rate prediction method
By monitoring government debt and labor productivity data in real time, extracting eigenvalues using K-mean clustering and hidden Markov model, and combining support vector machine models for exchange rate prediction, the problem of insufficient data integration and prediction accuracy in traditional methods is solved, and high-precision prediction of exchange rates and scientific understanding of economic trends is achieved.
Patent Information
- Application Number
- CN202510496387.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-21
- Publication Date
- 2025-08-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional exchange rate prediction methods rely on single or limited economic indicators, make it difficult to fully understand complex economic dynamics, lack effective data integration and high-quality preprocessing, and fail to make full use of machine learning technology, resulting in inaccurate prediction results.
By monitoring government debt and labor productivity data in real time, eigenvalues are extracted using K-mean clustering and hidden Markov model, exchange rate prediction is performed in combination with support vector machine model, and radial basis function and cross-validation are used to optimize model parameters.
A precise assessment of the country's fiscal health status and long-term economic growth potential has been achieved, which has significantly improved the accuracy and reliability of exchange rate forecasts and provided scientific decision-making support.
Smart Images
Figure CN120409794A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of financial data analysis and forecasting, and in particular to a method for exchange rate forecasting based on AI big data analysis. Background Art
[0002] Against the backdrop of deepening global economic integration, exchange rates, as a key factor connecting domestic and international markets, have a profound impact on international trade, investment decisions, and monetary policy formulation. Accurately forecasting exchange rate trends not only helps financial institutions optimize their portfolios and risk management strategies, but also provides governments with a scientific basis for formulating more effective macroeconomic policies. However, exchange rates are influenced by a variety of complex factors, including but not limited to government debt levels, labor productivity, inflation rates, interest rate differentials, and the international political and economic environment, making exchange rate forecasting a particularly challenging task.
[0003] The existing technology has the following deficiencies:
[0004] Traditional methods often rely on single or limited economic indicators for analysis, such as considering only the government debt ratio or a simple labor productivity growth rate, which leads to an incomplete and in-depth understanding of complex economic dynamics. In addition, traditional forecasting models (such as linear regression models) perform poorly when faced with nonlinear relationships and multivariate interactions, making it difficult to capture the deep patterns and trends hidden behind the data. Furthermore, existing technologies lack effective mechanisms to integrate data from different sources and perform high-quality preprocessing, resulting in inconsistencies and biases in the data input to the model, thereby affecting the accuracy of the final forecast results. Finally, traditional methods are also relatively lacking in parameter adjustment and model optimization, and are unable to fully utilize the powerful predictive capabilities of machine learning technology, limiting its effectiveness and adaptability in practical applications. These shortcomings make it difficult for existing technologies to meet the needs of modern economic analysis and decision support. Summary of the Invention
[0005] The purpose of the present invention is to provide a method for exchange rate prediction based on AI big data analysis to solve the problems in the above background.
[0006] The purpose of the present invention can be achieved through the following technical solutions:
[0007] A method for exchange rate prediction based on AI big data analysis, comprising the following steps:
[0008] S1: Real-time monitoring of government debt data and labor productivity data for the overall economy;
[0009] S2: Assess the current state of national finances based on monitored government debt data;
[0010] S3: Evaluate the long-term economic growth potential of a country based on the changing trend of labor productivity;
[0011] S4: Conduct a comprehensive analysis of the national fiscal health and the long-term economic growth potential of the country. Based on the analysis results, establish an exchange rate prediction model to accurately predict the future exchange rate.
[0012] As a further solution of the present invention: evaluating the current national fiscal health according to the monitored government debt data specifically includes:
[0013] Obtain the government debt data within the monitoring period. According to the government debt data, calculate the government debt ratio. According to the changing trend of the government debt ratio, calculate the eigenvalue of the government debt ratio. According to the eigenvalue of the government debt ratio, judge the current national fiscal health.
[0014] As a further solution of the present invention: the judgment of the current national fiscal health specifically includes:
[0015] Judge whether the eigenvalue of the government debt ratio of the current country is greater than or equal to the preset threshold. If so, the current national fiscal health is abnormal; if not, the current national fiscal health is normal.
[0016] As a further solution of the present invention: the process of obtaining the eigenvalue of the government debt ratio is as follows:
[0017] Obtain the government debt data at different time points within the monitoring period according to the time series, including the dataset of the government debt amounts at different time points within the monitoring period;
[0018] For each time point, calculate the government debt ratio by dividing the government debt amount at each time point by the gross domestic product in the same period.
[0019] After standardizing the calculated government debt ratio data, select the preset number of clusters, apply the K-means clustering algorithm to these data, initialize the K-means algorithm, allocate the data points to different clusters through iterative optimization, and calculate the new centroids of each cluster until convergence. In each identified cluster, calculate the geometric mean of the government debt ratio as the eigenvalue of the corresponding cluster, and calculate the mean value of the eigenvalues of all clusters to obtain the eigenvalue of the government debt ratio.
[0020] As a further solution of the present invention: evaluating the long-term economic growth potential of a country based on the changing trend of labor productivity specifically includes:
[0021] Obtain the labor productivity within the monitoring period. According to the changing trend of the labor productivity, calculate the change eigenvalue of the labor productivity. According to the change eigenvalue of the labor productivity, judge the long-term economic growth potential of the country.
[0022] As a further solution of the present invention: the judgment on whether the national economy will grow specifically includes:
[0023] Judging whether the eigenvalue of the change in the national labor productivity is greater than or equal to a preset threshold. If so, the national economy has no long-term economic growth potential; if not, the national economy has long-term economic growth potential.
[0024] As a further solution of the present invention: the process of obtaining the eigenvalue of the change in labor productivity is as follows:
[0025] According to the time series, obtain the labor productivity data at different time points within the monitoring period, and each time point corresponds to a specific labor productivity value;
[0026] Use the historical labor productivity data to train a hidden Markov model to estimate the initial state probability vector, state transition matrix, and observation probability distribution; the initial state probability vector represents the probability of each state at the start of the time series; the state transition matrix represents the probability of transitioning from one state to another; the observation probability distribution describes the likelihood of observing a specific labor productivity value in a particular state;
[0027] Apply the Viterbi algorithm to determine the state sequence that generates the observation sequence, and according to the obtained state sequence; the state sequence includes: the probability of transitioning from the recession state to the rapid growth state, the probability of transitioning from the rapid growth state to the recession state, and the probability of remaining stable; calculate the number of times of transitioning from the recession state to the rapid growth state divided by the total number of state transitions to obtain the eigenvalue of the change in labor productivity.
[0028] As a further solution of the present invention: the comprehensive analysis of the national fiscal health status and the long-term economic growth potential of the country, and according to the analysis results, establish an exchange rate prediction model, specifically including:
[0029] Obtain the eigenvalue of the government debt ratio of the country and the eigenvalue of the change in labor productivity, construct the government debt ratio eigenvalue and the eigenvalue of the change in labor productivity into a comprehensive feature vector as the input of the exchange rate prediction model, and use minimizing the error between the predicted future exchange rate and the actual future exchange rate as the training objective. According to the trained exchange rate prediction model, output the future exchange rate, and the exchange rate prediction model is a support vector machine model.
[0030] As a further solution of the present invention: the training process of the exchange rate prediction model is:
[0031] Obtain historical multiple sets of comprehensive feature vectors and exchange rate data to construct a training sample set, select the radial basis function as the kernel function to improve the fitting ability of the model; adjust the key parameters of the support vector machine model, including: penalty coefficient and kernel function parameters, and use the cross-validation method to select the best parameter combination; use the loss function to set the goal as minimizing the error between the predicted future exchange rate and the actual future exchange rate, and accurately predict the future exchange rate according to the trained model.
[0032] The beneficial effects of the present invention:
[0033] (1) By integrating real-time data collection and advanced data analysis techniques, the present invention achieves an accurate assessment of the national fiscal health status and long-term economic growth potential. Specifically, an automated system is used to regularly capture the latest government debt information and key indicators reflecting the labor market conditions from official sources, ensuring the timeliness and accuracy of the data. On this basis, the K-means clustering algorithm is used to perform clustering analysis on the standardized government debt ratio, identify different debt patterns, and calculate the geometric mean of the government debt ratio in each cluster to quantify the fiscal health status. At the same time, the hidden Markov model is applied to deeply explore the potential state transition laws of labor productivity over time, determine the most likely state sequence through the Viterbi algorithm, and extract the key state transition probabilities as the change characteristic values of labor productivity. This method not only significantly improves the accuracy of the assessment of the national fiscal health status and long-term economic growth potential, but also reveals potential economic trends and risks, providing scientific and reliable decision-making support for policymakers. In particular, it emphasizes the importance of the ability to recover from recession for long-term economic growth, and through in-depth analysis of various economic indicators, enhances the understanding of economic dynamic changes, enabling policymakers to respond more timely and effectively to potential economic challenges and promote the stable development of the economy. This comprehensive method represents a new paradigm of comprehensive and dynamic economic forecasting, greatly enhancing the reliability and practicality of the forecasting results.
[0034] (2) The present invention innovatively adopts a support vector machine model and combines the radial basis function as the kernel function to optimize the model performance by minimizing the error between the predicted future exchange rate and the actual future exchange rate, thereby achieving a highly accurate prediction of the future exchange rate. Specifically, this method first constructs a comprehensive feature vector, which consists of the eigenvalue of the government debt ratio and the eigenvalue of the change in labor productivity. These eigenvalues are calculated through advanced data analysis techniques such as the K-means clustering algorithm and the hidden Markov model. This multi-dimensional and in-depth data processing method not only fully considers the impacts of various key economic indicators but also effectively utilizes the powerful prediction ability of machine learning technology, significantly exceeding the limitations of traditional single-index prediction methods. In addition, the key parameters of the support vector machine model, including the penalty coefficient and the kernel function parameters, are carefully adjusted through the cross-validation method, further enhancing the adaptability and generalization ability of the model and ensuring its stability and reliability under different market conditions. For financial market participants and policymakers, this means that they can make more informed investment decisions or monetary policy adjustments based on more accurate and reliable exchange rate predictions, thereby effectively promoting economic stability and development. The present invention provides a comprehensive and dynamic economic prediction framework, greatly enhancing the understanding and control ability of global economic trends and providing a solid technical support for achieving more efficient resource allocation and risk management. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The present invention will be further described below with reference to the accompanying drawings.
[0036] Figure 1 It is a specific step flow block diagram of a method for predicting exchange rates based on AI big data analysis according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0037] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.
[0038] Please refer to Figure 1 As shown, the present invention is a method for predicting exchange rates based on AI big data analysis, including the following steps:
[0039] S1: Real-time monitor the government debt data and the labor productivity data of the overall economy;
[0040] S2: Evaluate the current national fiscal health status according to the monitored government debt data;
[0041] S3: Evaluate the long-term economic growth potential of a country through the changing trend of labor productivity;
[0042] S4: Conduct a comprehensive analysis of the national fiscal health and the long-term economic growth potential of the country. Based on the analysis results, establish an exchange rate forecasting model to accurately predict the future exchange rate.
[0043] In S1, the government debt data and the labor productivity data of the overall economy are monitored in real time, specifically including:
[0044] In the process of monitoring the government debt data in real time, first, an automated data collection system needs to be established. The system can regularly capture the latest government debt information from multiple official sources (such as the website of the Ministry of Finance). This process includes identifying and accessing reliable online resources that provide the latest government debt data, regularly extracting this data through API interfaces or web crawler technologies, and storing it in a central database for subsequent analysis. To ensure the accuracy and integrity of the data, the system also needs to have a data verification mechanism, such as methods for checking data consistency and handling missing values or outliers.
[0045] For the real-time monitoring of the labor productivity data of the overall economy, key indicators that can reflect the labor market conditions and changes in production efficiency need to be concerned, such as output per hour, unit labor cost, etc. This type of data can usually be obtained from the national statistical bureau, the central bank, and industry reports. The collection process involves setting up automated scripts or tools for regularly accessing these data sources and downloading the latest labor productivity statistics. In addition, these data need to be preprocessed, including but not limited to data cleaning, format conversion, and standardization, to ensure that data from different sources can be seamlessly integrated into the existing analysis framework to support further economic analysis and decision-making. The entire data collection process emphasizes timeliness and accuracy to ensure that policymakers can make judgments based on the latest and most accurate information.
[0046] In S2, based on the monitored government debt data, evaluate the current national fiscal health, specifically including:
[0047] Obtain the government debt data within the monitoring period, calculate the government debt ratio based on the government debt data, calculate the eigenvalue of the government debt ratio according to the changing trend of the government debt ratio, and judge the current national fiscal health based on the eigenvalue of the government debt ratio;
[0048] The judgment of the current national fiscal health specifically includes:
[0049] Judge whether the eigenvalue of the government debt ratio of the current country is greater than or equal to the preset threshold. If so, the current national fiscal health is abnormal; if not, the current national fiscal health is normal;
[0050] The process of obtaining the eigenvalue of the government debt ratio is as follows:
[0051] According to the time series, obtain the government debt data at different time points within the monitoring period, including the dataset of the government debt amounts at different time points within the monitoring period;
[0052] For each time point, calculate the government debt ratio by dividing the government debt amount at each time point by the gross domestic product in the same period;
[0053] After standardizing the calculated government debt ratio data, select the preset number of clusters, apply the K-means clustering algorithm to these data, initialize the K-means algorithm, allocate the data points to different clusters through iterative optimization, calculate the new centroids of each cluster until convergence, evaluate the clustering quality using the silhouette coefficient to ensure the rationality of the clustering results. In each identified cluster, calculate the geometric mean of the government debt ratio as the eigenvalue of the corresponding cluster, and calculate the mean value of the eigenvalues of all clusters to obtain the eigenvalue of the government debt ratio.
[0054] It should be noted that: In the present invention, by collecting the dataset of the government debt amounts at different time points within the monitoring period, then calculating the government debt ratio based on the ratio of the government debt amount at each time point to the gross domestic product in the same period, further using the K-means clustering algorithm to perform clustering analysis on the standardized government debt ratio to identify different debt patterns, quantifying the fiscal health by calculating the geometric mean of the government debt ratios in each cluster and comprehensively obtaining the eigenvalue of the government debt ratio. Finally, judge whether the national finance is healthy by comparing the eigenvalue of the government debt ratio with the preset threshold. This method not only improves the accuracy and efficiency of fiscal health assessment, but also can reveal potential economic trends and risks, providing a scientific decision-making basis for policymakers, reflecting the innovative application and technological progress in the field of fiscal data analysis. In addition, using the silhouette coefficient ensures the rationality of the clustering results and enhances the reliability of the model. This process combines real-time data collection, precise data processing, and effective pattern recognition technology, representing a new comprehensive and dynamic method for fiscal health assessment.
[0055] In S3, evaluate the long-term economic growth potential of the country through the change trend of labor productivity, specifically including:
[0056] Obtain the labor productivity within the monitoring period, calculate the eigenvalue of the labor productivity change according to the change trend of the labor productivity, and judge the long-term economic growth potential of the country according to the eigenvalue of the labor productivity change;
[0057] The judgment of whether the national economy will grow specifically includes:
[0058] Judge whether the eigenvalue of the change in the national labor productivity is greater than or equal to the preset threshold. If so, the national economy has no long-term economic growth potential. If not, the national economy has long-term economic growth potential;
[0059] The process of obtaining the eigenvalue of the change in labor productivity is as follows:
[0060] According to the time series, obtain the labor productivity data at different time points within the monitoring period, and each time point corresponds to a specific labor productivity value;
[0061] Use the historical labor productivity data to train the hidden Markov model to estimate the initial state probability vector, state transition matrix, and observation probability distribution; the initial state probability vector represents the probability of each state at the start of the time series; the state transition matrix represents the probability of transitioning from one state to another; the observation probability distribution describes the likelihood of observing a specific labor productivity value in a particular state;
[0062] Apply the Viterbi algorithm to determine the state sequence that generates the observation sequence, and according to the obtained state sequence; the state sequence includes: the probability of transitioning from the recession state to the rapid growth state, the probability of transitioning from the rapid growth state to the recession state, and the probability of remaining stable; calculate the number of times of transitioning from the recession state to the rapid growth state divided by the total number of state transitions to obtain the eigenvalue of the change in labor productivity.
[0063] It should be noted that: The system of the present invention systematically collects the labor productivity data at different time points within the monitoring period, and uses these data to train the hidden Markov model to estimate the initial state probability vector, state transition matrix, and observation probability distribution, so as to capture the potential patterns of the change of labor productivity over time. Apply the Viterbi algorithm to determine the most likely state sequence, and extract the information of key state transitions such as from recession to rapid growth from it. By calculating the probability of transitioning from the recession state to the rapid growth state as the eigenvalue of the change in labor productivity, the growth potential of the national economy is judged. This method can not only deeply explore the hidden trends and patterns behind labor productivity, but also provide a quantitative index for evaluating the health status of the national economy, reflecting the innovative application and technological progress in the field of economic forecasting. In particular, it emphasizes the importance of the ability to recover from recession for long-term economic growth, providing valuable insights and support for policymakers. This comprehensive way of using time series analysis and machine learning techniques makes the understanding of economic dynamic changes more accurate and improves the reliability and practicality of the prediction results.
[0064] In S4, comprehensively analyze the national fiscal health status and the long-term economic growth potential of the country, and establish an exchange rate prediction model based on the analysis results to accurately predict the future exchange rate, specifically including:
[0065] Obtain the eigenvalue of the government debt ratio of a country and the eigenvalue of the change in labor productivity. Construct the eigenvalue of the government debt ratio and the eigenvalue of the change in labor productivity into a comprehensive feature vector, which is used as the input of the exchange rate prediction model. Take minimizing the error between the predicted future exchange rate and the actual future exchange rate as the training objective. According to the trained exchange rate prediction model, output the future exchange rate. The exchange rate prediction model is a support vector machine model;
[0066] The training process of the exchange rate prediction model is as follows:
[0067] Obtain multiple historical sets of comprehensive feature vectors and exchange rate data to construct a training sample set. Select the radial basis function as the kernel function to improve the fitting ability of the model; adjust the key parameters of the support vector machine model, including: the penalty coefficient and the kernel function parameter. Use the cross-validation method to select the best parameter combination; use the loss function to set the goal of minimizing the error between the predicted future exchange rate and the actual future exchange rate. According to the trained model, accurately predict the future exchange rate.
[0068] It should be noted that: The innovation of the present invention lies in integrating the eigenvalue of the government debt ratio and the eigenvalue of the change in labor productivity, constructing a comprehensive feature vector as input data, and using a support vector machine model to accurately predict the future exchange rate. This method not only breaks through the limitations of traditional single economic indicator prediction, but also makes full use of the rich information provided by multi-source heterogeneous data, enhancing the model's ability to understand and capture complex market dynamics. Specifically, the radial basis function is used as the kernel function, and the key parameters of the model, such as the penalty coefficient and the kernel function parameter, are carefully adjusted through the cross-validation method, significantly improving the fitting ability and generalization performance of the model. In addition, using the loss function to set the goal of minimizing the prediction error ensures the high accuracy of future exchange rate prediction. This comprehensive analysis method combines the advantages of real-time data collection, advanced data analysis techniques, and machine learning algorithms, providing a comprehensive and dynamic new paradigm for exchange rate prediction, providing a more scientific and reliable decision-making basis for policymakers and financial market participants, and greatly promoting the effectiveness and accuracy of economic management and risk control. This innovative solution represents a frontier exploration in improving the accuracy and reliability of economic prediction in the context of big data and artificial intelligence.
[0069] Working principle of the present invention: The present invention monitors the government debt data and the labor productivity data of the overall economy in real time, and conducts in-depth analysis by combining these data to evaluate the national fiscal health status and the long-term economic growth potential, and finally establishes a support vector machine model to accurately predict the future exchange rate. First, in the data collection stage, the present invention uses an automated system to regularly capture the latest government debt information and key indicators reflecting the labor market conditions and productivity changes from multiple official sources, ensuring the timeliness and accuracy of the data. Then, the government debt ratio is calculated using the collected data, and the K-means clustering algorithm is applied to identify different debt patterns. At the same time, the hidden Markov model is used to analyze the change trend of labor productivity, and the government debt ratio eigenvalue and the labor productivity change eigenvalue are obtained respectively, which are used as important bases for measuring the national fiscal health status and economic growth potential. Subsequently, these two eigenvalues are constructed into a comprehensive feature vector and used as input data for training the SVM model. In the model training process, the radial basis function is selected as the kernel function, and by adjusting key parameters such as the penalty coefficient and the kernel function parameter, the cross-validation method is used to optimize the model performance, with the goal of minimizing the error between the predicted future exchange rate and the actual future exchange rate. This method not only improves the accuracy of exchange rate prediction, but also can reveal potential economic trends and risks, providing a scientific decision-making basis for policymakers. In particular, its innovative combination of real-time data collection, precise data processing, effective pattern recognition technology and advanced machine learning algorithms represents a new comprehensive and dynamic economic prediction method. This method emphasizes the importance of the ability to recover from recessions for long-term economic growth, and through in-depth analysis of a variety of economic indicators, it enhances the understanding of economic dynamic changes, thus achieving more accurate and reliable future exchange rate prediction. This process not only enhances the reliability and practicality of the model, but also provides strong support for understanding and predicting global economic changes.
[0070] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.
[0071] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, or a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.
[0072] It should be understood that the term "and / or" in this document is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood with reference to the context.
[0073] It should be understood that in various embodiments of the present application, the sequence numbers of the above processes do not indicate the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.
[0074] The above has described in detail an embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the present application should still fall within the scope covered by the present invention.
Claims
1. A method for exchange rate prediction based on AI big data analysis, characterized in that, It includes the following steps: S1: Monitor the government debt data and the labor productivity data of the overall economy in real time; S2: Evaluate the current national fiscal health status based on the monitored government debt data; S3: Evaluate the long-term economic growth potential of the country through the changing trend of labor productivity; S4: Conduct a comprehensive analysis of the national fiscal health status and the long-term economic growth potential of the country. According to the analysis results, establish an exchange rate prediction model to accurately predict the future exchange rate.
2. The method for exchange rate prediction based on AI big data analysis according to claim 1, wherein The evaluation of the current national fiscal health status based on the monitored government debt data specifically includes: Obtain the government debt data within the monitoring period. According to the government debt data, calculate the government debt ratio. According to the changing trend of the government debt ratio, calculate the eigenvalue of the government debt ratio. According to the eigenvalue of the government debt ratio, judge the current national fiscal health status.
3. The method for exchange rate prediction based on AI big data analysis according to claim 2, wherein, The judgment of the current national fiscal health status specifically includes: Judge whether the eigenvalue of the government debt ratio of the current country is greater than or equal to the preset threshold. If so, the current national fiscal health status is abnormal. If not, the current national fiscal health status is normal.
4. A method for exchange rate prediction based on AI big data analysis according to claim 3, characterized in that, The process of obtaining the eigenvalue of the government debt ratio is as follows: Obtain the government debt data at different time points within the monitoring period according to the time series, including the dataset of the government debt amounts at different time points within the monitoring period; For each time point, calculate the government debt ratio by dividing the government debt amount at each time point by the gross domestic product in the same period; After standardizing the calculated government debt ratio data, select the preset number of clusters, apply the K-means clustering algorithm to these data, initialize the K-means algorithm, allocate the data points to different clusters through iterative optimization, and calculate the new centroids of each cluster until convergence. In each identified cluster, calculate the geometric mean of the government debt ratio as the eigenvalue of the corresponding cluster, and calculate the mean of the eigenvalues of all clusters to obtain the eigenvalue of the government debt ratio.
5. A method for exchange rate prediction based on AI big data analysis according to claim 1, characterized in that, The evaluation of the long-term economic growth potential of the country through the changing trend of labor productivity specifically includes: Obtain the labor productivity within the monitoring period. According to the changing trend of the labor productivity, calculate the eigenvalue of the labor productivity change. According to the eigenvalue of the labor productivity change, judge the long-term economic growth potential of the country.
6. The method for predicting exchange rates based on AI big data analysis according to claim 5, characterized in that, The judgment of whether the national economy will grow specifically includes: Judge whether the eigenvalue of the labor productivity change of the country is greater than or equal to the preset threshold. If so, the country's economy has no long-term economic growth potential. If not, the country's economy has long-term economic growth potential.
7. A method for predicting exchange rates based on AI big data analysis according to claim 6, characterized in that, The process of obtaining the eigenvalue of the labor productivity change is as follows: Obtain the labor productivity data at different time points within the monitoring period according to the time series, and each time point corresponds to a specific labor productivity value; Use the historical labor productivity data to train the hidden Markov model to estimate the initial state probability vector, the state transition matrix, and the observation probability distribution; the initial state probability vector represents the probability of each state at the beginning of the time series; the state transition matrix represents the probability of transitioning from one state to another state; the observation probability distribution describes the possibility of observing a specific labor productivity value in a specific state; The Viterbi algorithm is applied to determine the state sequence that generates the observation sequence. According to the obtained state sequence, the state sequence includes: the probability of transitioning from a recession state to a rapid growth state, the probability of transitioning from a rapid growth state to a recession state, and the probability of remaining stable. Calculate the labor productivity change eigenvalue by dividing the number of times of transitioning from a recession state to a rapid growth state by the total number of state transitions.
8. A method for exchange rate prediction based on AI big data analysis according to claim 1, characterized in that, The comprehensive analysis of the national fiscal health and the long-term economic growth potential of the country is carried out. According to the analysis results, an exchange rate prediction model is established, which specifically includes: Obtain the government debt ratio eigenvalue and the labor productivity change eigenvalue of the country. Construct the government debt ratio eigenvalue and the labor productivity change eigenvalue into a comprehensive feature vector as the input of the exchange rate prediction model. Take minimizing the error between the predicted future exchange rate and the actual future exchange rate as the training objective. According to the trained exchange rate prediction model, output the future exchange rate. The exchange rate prediction model is a support vector machine model.
9. A method for exchange rate prediction based on AI big data analysis according to claim 8, characterized in that The training process of the exchange rate prediction model is as follows: Obtain multiple historical sets of comprehensive feature vectors and exchange rate data to construct a training sample set. Select the radial basis function as the kernel function to improve the fitting ability of the model. Adjust the key parameters of the support vector machine model, including the penalty coefficient and the kernel function parameter. Use the cross-validation method to select the best parameter combination. Use the loss function to set the goal of minimizing the error between the predicted future exchange rate and the actual future exchange rate. According to the trained model, accurately predict the future exchange rate.