Financial data denoising method based on time sequence support vector regression

Abstract

The invention discloses a financial data denoising method based on time sequence support vector regression, and mainly relates to the field of wavelet denoising. Comprising the following treatment steps. Preprocessing the original data; constructing a time series data set for the original data; dividing the implementation sequence data set into a training set and a test set; selecting a support vector regression method and a kernel function, training a support vector regression model and adjusting parameters; selecting the trained support vector regression model to carry out regression prediction on the data, and identifying and filtering noise in the data; and optimizing the support vector regression model. The time sequence support vector regression method has the beneficial effects that the time sequence support vector regression method is suitable for processing complex financial time sequence data. The SVR method is used for modeling and predicting financial data, rules and trends in the data can be effectively captured, and the accuracy and reliability of data analysis are improved.

Description

Technical Field

[0001] The present invention relates to the field of wavelet denoising, and specifically to a financial data denoising method based on time series support vector regression. Background Art

[0002] The statements in this part merely provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] Due to the influence of various accidental factors in the financial market, there are many noises in financial data, especially financial time series data. These noises may have a negative impact on data analysis and prediction results. Therefore, how to effectively remove the noises in financial data and improve the accuracy and reliability of the data has become an important issue in financial data processing.

[0004] Traditional methods for denoising financial time series mainly include the moving average method, traditional filtering methods, Kalman filtering, and Wiener filtering methods; while the moving average method is only applicable to simple data processing during the denoising process. The traditional filtering method uses Fourier transform to regard high-frequency signals as noises and sets all Fourier coefficients higher than a certain threshold frequency to zero, thereby achieving denoising. Kalman filtering requires knowing the motion law of the system to establish an accurate equation. However, financial time series is a non-stationary and non-linear time series, and it is difficult to describe it with a definite equation. The Wiener filtering method is only applicable to stationary processes and requires prior knowledge of noises and useful signals. Therefore, a method of time series support vector regression (SVR) is adopted here to process complex financial time series data. Using the SVR method to model and predict financial data can effectively capture the laws and trends in the data and improve the accuracy and reliability of data analysis.

[0005] For the method of time series support vector regression, denoising financial data can improve the accuracy and reliability of the data and effectively improve the results of financial data analysis. Summary of the Invention

[0006] The purpose of the present invention is to provide a financial data denoising method based on time series support vector regression, which is applicable to processing complex financial time series data. Using the SVR method to model and predict financial data can effectively capture the laws and trends in the data and improve the accuracy and reliability of data analysis. By establishing an SVR model and performing regression analysis on financial data, regression prediction of financial data can be effectively carried out, noises in the data can be identified and filtered out, and the quality and reliability of the data can be improved.

[0007] To achieve the above object, the present invention is realized through the following technical solutions:

[0008] A financial data denoising method based on time series support vector regression includes the following processing steps;

[0009] S1. Preprocess the original data;

[0010] S2. Construct a time series data set from the original data;

[0011] S3. Divide the implementation sequence data set into a training set and a test set;

[0012] S4. Select the support vector regression method and kernel function, train the support vector regression model and adjust the parameters;

[0013] S5. Select the trained support vector regression model to perform regression prediction on the data, and identify and filter the noise in the data;

[0014] S6. Optimize the support vector regression model;

[0015] In step S1, the preprocessing adopts the mean method, and the mean value of the attribute values is used to fill the vacancies in the original data. The attribute values include the opening price, highest price, lowest price, closing price, daily return rate, trading volume, trading amount, and trading time of a single day.

[0016] Specifically, for the time series data set, interpolation, filling, or deletion of the missing values is performed at the positions of the missing attribute values to ensure the integrity of the data.

[0017] In step S3, the training set and the test set are divided by time. The earlier data is used as the training set, and the later data is used as the test set.

[0018] The training set learns the historical data patterns and trends to predict future data, and compares and evaluates with the data on the test set. The training set uses the cross-validation method to select the best model parameters.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0020] The method of time series support vector regression is applicable to processing complex financial time series data. Using the SVR method to model and predict financial data can effectively capture the laws and trends in the data, and improve the accuracy and reliability of data analysis. By establishing an SVR model and performing regression analysis on financial data, regression prediction of financial data can be effectively carried out, the noise in the data can be identified and filtered out, and the quality and reliability of the data can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Attached Figure 1 is a step diagram of a financial data denoising method based on time series support vector regression of the present invention.

[0022] Appendix Figure 2 This is the denoising effect diagram of a financial data denoising method based on time series support vector regression of the present invention. Specific implementation manners

[0023] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by this application.

[0024] The present invention relates to a financial data denoising method based on time series support vector regression.

[0025] The implementation steps of the financial data denoising method include a data preparation stage, a support vector regression (SVR) model training stage, a data denoising stage, a model evaluation and optimization stage, and a data analysis and application stage.

[0026] Data preparation stage; it is necessary to collect financial time series data, including information such as stock prices, trading volumes, opening prices, highest prices, lowest prices, closing prices, daily returns, trading volumes, trading amounts, and trading times on a single day. And clean and preprocess the data, handle missing values and outliers to ensure the quality of the data. For outliers, methods such as deletion are selected for processing, and for the positions of missing values and the deleted outliers, interpolation and filling methods can be used to fill the missing positions, or the values at the missing positions can be selected for deletion to ensure the accuracy and integrity of the data and ensure the quality of the data.

[0027] Support vector regression (SVR) model training stage; divide the data set into a training set and a test set. Usually, a ratio of 70% training set and 30% test set is used for future prediction. The training set continuously learns the historical data patterns and trends to predict future data, and compares and evaluates the data on the test set to optimize the parameters of the training set and select the best model parameters to improve the performance and generalization ability of the model. The distinction between the training set and the test set is that the data with an earlier time is used as the training set, and the later data is used as the test set. The model of the training set can be optimized through cross-validation methods and kernel functions (linear kernel, polynomial kernel, radial basis kernel, etc.) so that the training set trains the support vector regression (SVR) model and adjusts it to the optimal state.

[0028] Data denoising processing stage; First, it is necessary to ensure that the format and features of the data are consistent with the data used during model training. Then, use the trained Support Vector Regression (SVR) model to perform regression prediction on the original data to obtain the fitting values of the model for the data. Next, compare the original data with the model fitting values, identify and filter out the noise in the data, and obtain the denoised data.

[0029] Model evaluation and optimization stage; Evaluate the Support Vector Regression (SVR) model through methods such as mean squared error, R-squared, cross-validation, and learning curves, calculate the regression error metrics, adjust the model according to the evaluation results, optimize the model parameters, control the complexity of the model through regularization to prevent overfitting, and attempt to integrate multiple SVR models to improve the generalization ability of the model. Furthermore, optimize and adjust the model to improve accuracy and generalization ability.

[0030] Data analysis and application stage; Use the denoised data for further data analysis, modeling, and prediction. Make financial decisions based on more accurate data to improve the accuracy and reliability of data analysis. During the data analysis process, exploratory data analysis can be performed on the denoised data, including statistical description, data visualization, etc., to understand the distribution of the data and the relationships between features. Feature selection is to select features that have an important impact on the target variable, and use feature importance evaluation methods to assist in the selection. The data processing of the model selects appropriate models according to business requirements, such as linear regression, support vector regression, cross-validation, etc. And when training the model, select the denoised model to ensure the quality of the data. After training, use the model to predict new data, and make corresponding decisions or actions based on the prediction results. Then regularly monitor the performance of the model and optimize the model to ensure the stability of the model.

[0031] Detailed explanation of usage:

[0032] Preprocess the original data;

[0033] Construct a time series dataset from the original data;

[0034] Divide the time series dataset into a training set and a test set;

[0035] Select the support vector regression method and kernel function, train the support vector regression model and adjust the parameters;

[0036] Select the trained support vector regression model to perform regression prediction on the data, identify and filter the noise in the data;

[0037] Optimize the support vector regression model.

[0038] In summary, the time series support vector regression method is applicable to processing complex financial time series data. Using the SVR method to model and predict financial data can effectively capture the patterns and trends in the data, improving the accuracy and reliability of data analysis. By establishing an SVR model and performing regression analysis on financial data, regression prediction of financial data can be effectively carried out, identifying and filtering out the noise in the data, and improving the quality and reliability of the data.

Claims

1. A financial data denoising method based on time series support vector regression, characterized in that: It includes the following processing steps; S1. Preprocess the original data; S2. Construct a time series data set from the original data; S3. Divide the implementation sequence data set into a training set and a test set; S4. Select a support vector regression method and a kernel function, train a support vector regression model and adjust the parameters; S5. Select the trained support vector regression model to perform regression prediction on the data, and identify and filter the noise in the data; S6. Optimize the support vector regression model.

2. The financial data denoising method based on time series support vector regression according to claim 1, characterized in that: In step S1, the preprocessing adopts the mean method, and the mean value of the attribute values is used to fill the vacancies in the original data.

3. The financial data denoising method based on time series support vector regression according to claim 2, characterized in that: The attribute values include the opening price, highest price, lowest price, closing price, daily return rate, trading volume, transaction amount, and trading time of a single day.

4. The financial data denoising method based on time series support vector regression according to claim 3, characterized in that: Specifically, for the time series data set, interpolation, filling, or deletion of missing values is used at the positions of missing attribute values to ensure data integrity.

5. The financial data denoising method based on time series support vector regression according to claim 4, characterized in that: In step S3, the training set and the test set are divided by time. The earlier data is used as the training set, and the later data is used as the test set.

6. The financial data denoising method based on time series support vector regression according to claim 5, wherein: The training set learns the patterns and trends of historical data to predict future data, and compares and evaluates it with the data on the test set.

7. A financial data denoising method based on time series support vector regression according to claim 6, characterized in that: The training set uses the cross-validation method to select the best model parameters.

Images