Blood glucose concentration prediction method based on mixed kernel function Gaussian process regression and electronic equipment

Abstract

The invention provides a blood glucose concentration prediction method based on mixed kernel function Gaussian process regression and electronic equipment. The method comprises the following steps: firstly, acquiring historical blood glucose concentration data of a user and time sequence data related to blood glucose change, and preprocessing the acquired data; then feature extraction is carried out on the basis of the preprocessed data, feature vectors are constructed, corresponding blood glucose concentrations serve as labels, and therefore a sample set is constructed; then establishing a nonlinear relationship between the feature vectors and the tags by adopting a mixed kernel function Gaussian process regression model, and optimizing parameters of the blood glucose concentration prediction model by utilizing a swarm intelligence algorithm to obtain an optimal blood glucose concentration prediction model; and inputting the current feature vector into an optimal blood glucose concentration prediction model, and taking a corresponding prediction label output by the optimal blood glucose concentration prediction model as a blood glucose concentration prediction result. The accuracy of blood glucose concentration prediction can be improved.

Description

Technical Field

[0001] This application belongs to the field of blood glucose concentration prediction technology, specifically relating to a blood glucose concentration prediction method and electronic device based on mixed kernel function Gaussian process regression. Background Technology

[0002] Accurate prediction of blood glucose concentration is a core technology in intelligent diabetes management, directly impacting clinical decisions such as insulin dosage adjustment and hypoglycemia warning. Currently, blood glucose prediction methods are mainly divided into two categories: one is computational methods based on physiological models, which predict blood glucose levels by simulating the physiological mechanisms involved in human metabolism; the other is data-driven methods based on machine learning, which use a large amount of historical data to train models for prediction.

[0003] However, existing methods for predicting blood glucose concentration still have significant limitations, and their accuracy needs to be improved. Summary of the Invention

[0004] The purpose of this application is to provide a blood glucose concentration prediction method and electronic device based on mixed kernel function Gaussian process regression, which can improve the accuracy of blood glucose concentration prediction.

[0005] The technical solution adopted in this application is: A method for predicting blood glucose concentration based on Gaussian process regression with mixed kernel functions includes: Acquire users' historical blood glucose concentration data and time series data related to blood glucose changes, and preprocess the acquired data; Feature extraction is performed on the preprocessed data to construct feature vectors, and the corresponding blood glucose concentration is used as a label to construct a training sample set. A blood glucose concentration prediction model is established using a Gaussian process regression model with a mixture kernel function. This model describes the nonlinear relationship between the feature vector and the label. Specifically, the model calculates the covariance between the current feature vector and the feature vector of each training sample using a mixture kernel function, and combines the contribution weight of each training sample to the prediction result to aggregate the information of all training samples to generate the predicted label corresponding to the current feature vector. The contribution weight of each training sample to the prediction result is equal to the actual observed covariance matrix. The elements at the corresponding positions are obtained by multiplying the inverse matrix of the training sample set by the vector composed of the labels of all training samples. Based on the training sample set and the hybrid kernel function, the covariance matrix is ​​constructed as follows: First, the covariance between each pair of feature vectors in the training sample set is calculated using the hybrid kernel function; then, the covariance matrix between feature vectors in the training sample set is constructed based on the calculated covariance. ; and then Adding noise to obtain ; Among them, the mixed kernel functions include the Gaussian kernel function and the rational quadratic kernel function; Based on the training sample set, a swarm intelligence algorithm is used to optimize the parameters of the blood glucose concentration prediction model to obtain the optimal blood glucose concentration prediction model. The current feature vector is input into the optimal blood glucose concentration prediction model, and the corresponding prediction label output by the optimal blood glucose concentration prediction model is used as the prediction result of blood glucose concentration.

[0006] In one possible implementation, time-series data related to blood glucose changes include: insulin injection dose / time, carbohydrate intake / time, exercise type / intensity / time, and heart rate.

[0007] In one possible implementation, the acquired data is preprocessed, including: handling missing values, detecting and correcting outliers, time alignment, and normalization.

[0008] In one possible implementation, feature extraction is performed based on the preprocessed data to construct a feature vector, including: Features are extracted from historical blood glucose concentration data, including: rate of change of blood glucose, acceleration of change of blood glucose, mean, variance and extreme values ​​of blood glucose over a period of time; Features were extracted from time-series data related to blood glucose changes, including: insulin activity curve estimation, carbohydrate absorption curve estimation, and quantitative indicators of exercise intensity. Feature vectors are constructed based on the extracted features.

[0009] In one possible implementation, let the training sample set be... ; For the first The feature vectors of each training sample For the first The true labels of each training sample The number of training samples; The blood glucose concentration prediction model is as follows: ; in, For the current feature vector Corresponding predicted labels; express and The covariance between them is calculated based on the mixture kernel function; ; Indicates the first The contribution weights of each training sample to the prediction result are set to a value of . The One element; The formula for calculating the covariance between pairwise feature vectors of the training sample set based on the hybrid kernel function is: ; in: express and Covariance between For the first The feature vectors of each training sample ; and The weights and hyperparameters of the Gaussian kernel function and the rational quadratic kernel function are respectively. To control the amplitude of the Gaussian kernel, hyperparameters This is the scale factor for distance calculation; These are shape parameters that control the weighting between different scales; This is a length scale parameter that controls the smoothness of the changes; , , , , , All were obtained through optimization using swarm intelligence algorithms; Construct the covariance matrix between the feature vectors of the training sample set based on the calculated covariance. The formula is: ; exist Adding noise to obtain The formula is: In the formula, for The identity matrix, The variance is Gaussian white noise; , This represents a vector consisting of the labels of all training samples in the training sample set.

[0010] In one possible implementation, the swarm intelligence algorithm employs the Fishing Optimization Algorithm (CFOA algorithm).

[0011] In one possible implementation, based on the training sample set, a fishing optimization algorithm is used to optimize the parameters of the blood glucose concentration prediction model to obtain the optimal blood glucose concentration prediction model, including: 1) Set the parameters of the fishing optimization algorithm, including: number of fishermen, maximum number of iterations, learning factor and cooperation factor; initialize the position of each fisherman; the position of each fisherman corresponds to a set of parameter values ​​for the blood glucose concentration prediction model; 2) Substitute the current location of each fisherman into the blood glucose concentration prediction model, and input the feature vector of each training sample as the current feature vector into the blood glucose concentration prediction model to obtain the corresponding predicted label; calculate the fitness value of each fisherman's current location based on the true label and predicted label of the training sample: ; In the formula For fitness value, For the first training samples The corresponding predicted label, For the first The true label of each sample; 3) Obtain the optimal fisherman's location and the best fitness value; 4) Based on the optimal individual and the best fitness value, update the position of each fisherman by combining the learning factor and the cooperation factor, and return to step 2) to iterate and optimize until the maximum number of iterations is reached. If so, output the parameter values ​​of the blood glucose concentration prediction model corresponding to the global optimal fisherman position and the global optimal fitness value as the optimal parameter values; substitute the optimal parameter values ​​into the blood glucose concentration prediction model to obtain the optimal blood glucose concentration prediction model.

[0012] In a second aspect, this application provides an electronic device, including: a memory and a processor; The memory is used to store computer programs; The processor is used to invoke the computer program to execute the method described above.

[0013] Thirdly, this application provides a computer-readable storage medium storing a computer program that, when run on an electronic device, causes the electronic device to perform the method described above.

[0014] Fourthly, this application provides a computer program product, including a computer program that, when run on an electronic device, causes the electronic device to perform the method described above.

[0015] The specific implementation methods of the second to fourth aspects of this application can refer to the implementation methods of the first aspect, and will not be elaborated here.

[0016] Beneficial effects: This application provides a blood glucose concentration prediction method and electronic device based on Gaussian Process Regression (GPR) with hybrid kernel functions. It effectively integrates the advantages of Gaussian kernel functions in capturing local fine changes and rational quadratic kernel functions in characterizing multi-scale features. For example, Gaussian kernel functions can capture local sensitive features such as postprandial blood glucose spikes, while rational quadratic kernel functions can fit the long-term trend of slow changes in basal metabolism, significantly enhancing the ability of Gaussian process regression models to represent complex nonlinear dynamics of blood glucose. In addition, a swarm intelligence algorithm is introduced for global automated optimization, which greatly improves model performance and overcomes the limitations and inefficiencies of manual parameter tuning. This significantly improves the accuracy and generalization ability of the blood glucose concentration prediction model, providing more reliable technical support for personalized blood glucose management for diabetic patients. Attached Figure Description

[0017] Figure 1 This is a flowchart of the application process. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present application, the technical solution of the present application will be further described in detail below with reference to the embodiments and accompanying drawings.

[0019] This application provides a blood glucose concentration prediction method and electronic device based on Gaussian Process Regression (GPR) with mixed kernel functions, offering more reliable technical support for personalized blood glucose management in diabetic patients. First, historical blood glucose concentration data and time-series data related to blood glucose changes are acquired and preprocessed. Then, feature extraction is performed on the preprocessed data to construct feature vectors, with the corresponding blood glucose concentrations serving as labels, thus building a sample set. Next, a nonlinear relationship between the feature vectors and labels is established using a Gaussian Process Regression model with mixed kernel functions. A swarm intelligence algorithm is then used to optimize the parameters of the blood glucose concentration prediction model, obtaining the optimal blood glucose concentration prediction model. The current feature vector is input into the optimal blood glucose concentration prediction model, and the corresponding prediction label output by the optimal model is used as the predicted blood glucose concentration.

[0020] The following will refer to Figure 1 A specific implementation method according to this application is described.

[0021] Example 1: This application provides a method for predicting blood glucose concentration based on Gaussian process regression with mixed kernel functions, including: S1. Obtain the user's historical blood glucose concentration data and time series data related to blood glucose changes, and preprocess the obtained data.

[0022] In some embodiments, time-series data related to changes in blood glucose include insulin injection dose / time, carbohydrate intake / time, exercise type / intensity / time, and heart rate.

[0023] In this embodiment, the character " / " indicates that the objects before and after it are in an "and" relationship. For example, A / B means: A and B.

[0024] The acquired data is then preprocessed, including handling missing values, outlier detection and correction, time alignment, and normalization.

[0025] S2. Based on the preprocessed data, feature extraction is performed to construct feature vectors, and the corresponding blood glucose concentration is used as a label to construct a sample set.

[0026] In some embodiments, features are extracted from historical blood glucose concentration data, including: blood glucose rate of change, blood glucose acceleration, mean blood glucose over a past period, variance, and extreme values; Features were extracted from time-series data related to blood glucose changes, including: insulin activity curve estimation, carbohydrate absorption curve estimation, and quantitative indicators of exercise intensity. Based on the features extracted above, a feature vector is constructed, and the corresponding blood glucose concentration is used as a label.

[0027] Considering that blood glucose changes are influenced by multiple factors such as insulin activity, carbohydrate absorption, and exercise intensity, this application extracts these features to construct feature vectors, trains the model, and makes predictions, which can improve prediction performance.

[0028] In this embodiment of the application, the sample set is divided into a training sample set and a test sample set. For example, the sample set and the test sample set are divided in an 8:2 ratio.

[0029] The training sample set can be described as The test sample set can be described as follows: ; For the first The feature vectors of each training sample ; For the first The feature vector of each test sample ; For the first The true labels of each training sample; For the first The true label of each test sample. and These represent the number of training and testing samples, respectively. , , This represents a vector composed of the feature vectors of all training samples in the training sample set. This represents a vector consisting of the true labels of all training samples in the training sample set; , , This represents a vector composed of the feature vectors of all test samples in the test sample set. It is a vector consisting of the true labels of all test samples in the test sample set.

[0030] S3. A blood glucose concentration prediction model is established using a Gaussian process regression model with mixed kernel functions; the blood glucose concentration prediction model is used to describe the nonlinear relationship between the feature vector and the label (blood glucose concentration); wherein, the mixed kernel function includes a Gaussian kernel function and a rational quadratic kernel function.

[0031] Given that blood glucose changes are influenced by multiple factors, including insulin activity, carbohydrate absorption, and exercise intensity, exhibiting complex characteristics of both rapid local fluctuations (such as postprandial peaks) and gradual long-term changes (such as basal metabolism), a single kernel function (such as a Gaussian kernel) is insufficient to simultaneously capture these multi-scale spatiotemporal features, leading to limited prediction accuracy. Existing methods lack refinement in extracting and fusing dynamic physiological features such as insulin activity curves, carbohydrate absorption rates, and exercise intensity quantification, limiting the model's ability to analyze blood glucose-driven mechanisms. Gaussian process regression (GPR), however, offers advantages such as quantifying prediction uncertainty and strong modeling capabilities for nonlinear relationships. Therefore, this application employs hybrid kernel function Gaussian process regression for modeling to enhance the model's expressive power.

[0032] The obtained feature vectors are used as inputs for model training, and the labels are used as outputs of the model.

[0033] In some embodiments, this step specifically includes: S3.1 Construct a hybrid kernel function, with the following expression: ; in: express and The covariance between two feature vectors can quantify the degree of similarity between them. and The weights and hyperparameters of the Gaussian kernel function and the rational quadratic kernel function are respectively. To control the amplitude of the Gaussian kernel, hyperparameters This is the scale factor for distance calculation; These are shape parameters that control the weighting between different scales; This is a length scale parameter that controls the smoothness of the changes.

[0034] S3.2 Calculate the covariance matrix between feature vectors of the training sample set using the constructed hybrid kernel function. : ; in, For an ideal noise-free covariance matrix, The feature vectors of the training sample set are calculated and used to construct the prior distribution of the Gaussian process.

[0035] Considering the impact of noise, in Adding noise, we obtain the actual observed covariance matrix. ,as follows: ; in, for The identity matrix, The variance of Gaussian white noise during Gaussian process regression is given.

[0036] S3.3 Constructing a blood glucose concentration prediction model: ; in, For the current feature vector Corresponding predicted labels; express and The covariance between them is calculated based on the mixture kernel function; Indicates the first The contribution weights of each training sample to the prediction result are set to a value of . The Each element.

[0037] S4. Based on the training sample set, the parameters of the blood glucose concentration prediction model are optimized using a swarm intelligence algorithm to obtain the optimal blood glucose concentration prediction model.

[0038] Considering the large number and strong coupling of hyperparameters in hybrid kernel functions, traditional optimization methods suffer from the curse of dimensionality, are prone to getting trapped in local optima, and have huge computational costs, making it difficult to meet the real-time requirements of blood glucose prediction. This application employs a swarm intelligence algorithm to optimize the parameters of the blood glucose concentration prediction model, thereby improving the parameter optimization effect.

[0039] In some embodiments, the swarm intelligence algorithm employs algorithms such as the Fishing Optimization Algorithm (CFOA), the Ant Colony Algorithm (ACO), and the Particle Swarm Optimization Algorithm (PSO).

[0040] In some embodiments, this step specifically includes: S4.1 Set the parameters of the fishing optimization algorithm, including: number of fishermen, maximum number of iterations, learning factor and cooperation factor; initialize the position of each fisherman; the position of each fisherman corresponds to a set of parameter values ​​for the blood glucose concentration prediction model; S4.2. Substitute the current position of each fisherman into the blood glucose concentration prediction model, and input the feature vector of each training sample as the current feature vector into the blood glucose concentration prediction model to obtain the corresponding predicted label; calculate the fitness value of each fisherman's current position based on the true label and predicted label of the training sample: ; In the formula For fitness value, For the first training samples The corresponding predicted label, For the first The true label of each sample; S4.3 Obtain the optimal fisherman's location and the best fitness value; S4.4. Based on the optimal individual and the best fitness value, update the position of each fisherman by combining the learning factor and the cooperation factor, and return to step S4.2 to iterate and optimize until the maximum number of iterations is reached. If so, output the parameter values ​​of the blood glucose concentration prediction model corresponding to the global optimal fisherman position and the global optimal fitness value as the optimal parameter values; substitute the optimal parameter values ​​into the blood glucose concentration prediction model to obtain the optimal blood glucose concentration prediction model.

[0041] After obtaining the optimal blood glucose concentration prediction model, the performance of the optimal blood glucose concentration prediction model can be tested using the training sample set.

[0042] During the testing phase: First, the covariance matrix between the feature vectors of the training and test sample sets is calculated using the constructed hybrid kernel function. : ; It is The matrix. Used for posterior prediction, the calculation calls the parameters of the already trained model. .

[0043] Then, the predicted value is output based on the optimal blood glucose concentration prediction model; ; In the formula, It is a vector consisting of the mean (optimal estimate) of the predicted labels of all test samples in the test sample set.

[0044] Next, we will assess the uncertainty of the forecast.

[0045] First, the covariance matrix between the feature vectors of the test sample set is calculated using the constructed hybrid kernel function. , It is Matrix; ;

[0046] It is The matrix.

[0047] Reconstruct the true label vectors of the training samples and the predicted label vector of the test sample The joint prior distribution is as follows: ; In the formula With a mean of 0 and a covariance of The normal distribution.

[0048] Through the By defining a Gaussian joint prior distribution, the predicted label vector can be analytically derived. The posterior distribution of can be written as ;in Let be a vector consisting of the mean (optimal estimate) of the predicted labels of all test samples in the test sample set. The covariance of the predicted labels for all test samples in the test sample set is calculated as follows: , ; Therefore, the model's prediction error and uncertainty can be assessed.

[0049] For example, the root mean square error (RMSE) is used to assess prediction error, and the calculation formula is as follows: ; In the formula, express The Middle The nth element represents the nth element in the test sample set. The mean of the predicted labels for each test sample.

[0050] The uncertainty of the model is assessed based on the coverage ratio, and the calculation formula is as follows: ; in, yes The Middle The diagonal element represents the nth element in the test sample set. The prediction standard deviation of each test sample; For indicator functions, If the internal condition is met, count 1; otherwise, do not count. 1.96 corresponds to the 95% confidence interval of a normal distribution. Calculate the proportion of samples that meet the condition.

[0051] The coverage ratio is calculated using a formula to determine the proportion of true labels in the test sample set that fall within the predicted confidence interval. If the coverage ratio is greater than a coverage threshold (e.g., 95%), it indicates that the model has low uncertainty and high reliability. Conversely, if the coverage ratio is less than the coverage threshold, it indicates that the parameters of the mixing kernel function or the noise term need to be adjusted. .

[0052] S5. Input the current feature vector into the optimal blood glucose concentration prediction model, and use the corresponding prediction label output by the optimal blood glucose concentration prediction model as the prediction result of blood glucose concentration.

[0053] This application effectively integrates the advantages of Gaussian kernel functions in capturing local fine changes and rational quadratic kernel functions in characterizing multi-scale features, significantly enhancing the ability of Gaussian process regression models to represent complex nonlinear dynamics of blood glucose. In addition, the introduction of swarm intelligence algorithms for global automated optimization greatly improves model performance and overcomes the limitations and inefficiencies of manual parameter tuning, significantly improving the accuracy and generalization ability of blood glucose concentration prediction models, and providing more reliable technical support for personalized blood glucose management for diabetic patients.

[0054] It should be understood that the above numbers S1~S5, S3.1~S3.3, and S4.1~S4.4 are only used to distinguish and facilitate the expression of different steps, and do not necessarily constitute a restriction on the execution order between the steps.

[0055] Example 2: This embodiment provides an electronic device, including: a memory and a processor; The memory is used to store computer programs; The processor is configured to invoke the computer program to execute the method as described in Embodiment 1.

[0056] Example 3: This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is run on an electronic device, it causes the electronic device to perform the method described in Embodiment 1.

[0057] Example 4: This embodiment provides a computer program product, including a computer program that, when run on an electronic device, causes the electronic device to perform the method described in Embodiment 1.

[0058] The specific implementation of the system, electronic device, computer-readable storage medium, and computer program product provided in this application can be referred to the specific embodiments of the above methods, and will not be repeated here.

[0059] The technical content of the above embodiments can be referred to each other. For the same or similar technical features, appropriate omissions have been made in some embodiments to avoid repeated descriptions.

[0060] Obviously, those skilled in the art should understand that the various units or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0061] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for predicting blood glucose concentration based on Gaussian process regression with mixed kernel functions, characterized in that, include: Acquire users' historical blood glucose concentration data and time series data related to blood glucose changes, and preprocess the acquired data; Feature extraction is performed on the preprocessed data to construct feature vectors, and the corresponding blood glucose concentration is used as a label to construct a training sample set. A blood glucose concentration prediction model is established using a Gaussian process regression model with a mixture kernel function. This model describes the nonlinear relationship between the feature vector and the label. Specifically, the model calculates the covariance between the current feature vector and the feature vector of each training sample using a mixture kernel function, and combines the contribution weight of each training sample to the prediction result to aggregate the information of all training samples to generate the predicted label corresponding to the current feature vector. The contribution weight of each training sample to the prediction result is equal to the actual observed covariance matrix. The elements at the corresponding positions are obtained by multiplying the inverse matrix of the training sample set by the vector composed of the labels of all training samples. Based on the training sample set and the hybrid kernel function, the covariance matrix is ​​constructed as follows: First, the covariance between each pair of feature vectors in the training sample set is calculated using the hybrid kernel function; then, the covariance matrix between feature vectors in the training sample set is constructed based on the calculated covariance. ; and then Adding noise to obtain ; Among them, the mixed kernel functions include the Gaussian kernel function and the rational quadratic kernel function; Based on the training sample set, a swarm intelligence algorithm is used to optimize the parameters of the blood glucose concentration prediction model to obtain the optimal blood glucose concentration prediction model. The current feature vector is input into the optimal blood glucose concentration prediction model, and the corresponding prediction label output by the optimal blood glucose concentration prediction model is used as the prediction result of blood glucose concentration.

2. The method according to claim 1, characterized in that, Time-series data related to blood glucose changes include: insulin injection dose / time, carbohydrate intake / time, exercise type / intensity / time, and heart rate.

3. The method according to claim 2, characterized in that, The acquired data is preprocessed, including: handling missing values, detecting and correcting outliers, time alignment, and normalization.

4. The method according to claim 2, characterized in that, Feature extraction is performed on the preprocessed data to construct feature vectors, including: Features are extracted from historical blood glucose concentration data, including: rate of change of blood glucose, acceleration of change of blood glucose, mean, variance and extreme values ​​of blood glucose over a period of time; Features were extracted from time-series data related to blood glucose changes, including: insulin activity curve estimation, carbohydrate absorption curve estimation, and quantitative indicators of exercise intensity. Feature vectors are constructed based on the extracted features.

5. The method according to any one of claims 1 to 4, characterized in that, Let the training sample set be ; For the first The feature vectors of each training sample For the first The true labels of each training sample The number of training samples; The blood glucose concentration prediction model is as follows: ； in, For the current feature vector Corresponding predicted labels; express and The covariance between them is calculated based on the mixture kernel function; Indicates the first The contribution weights of each training sample to the prediction result are set to a value of . The One element; The formula for calculating the covariance between pairwise feature vectors of the training sample set based on the hybrid kernel function is: ； in: express and Covariance between For the first The feature vectors of each training sample ; and The weights and hyperparameters of the Gaussian kernel function and the rational quadratic kernel function are respectively. To control the amplitude of the Gaussian kernel, hyperparameters This is the scale factor for distance calculation; These are shape parameters that control the weighting between different scales; This is a length scale parameter that controls the smoothness of the changes; , , , , , All were obtained through optimization using swarm intelligence algorithms; Construct the covariance matrix between the feature vectors of the training sample set based on the calculated covariance. The formula is: ； exist Adding noise to obtain The formula is: In the formula, for The identity matrix, The variance is Gaussian white noise; , This represents a vector consisting of the labels of all training samples in the training sample set.

6. The method according to claim 5, characterized in that, The swarm intelligence algorithm employs a fishing optimization algorithm.

7. The method according to claim 6, characterized in that, Based on the training sample set, the parameters of the blood glucose concentration prediction model are optimized using a fishing optimization algorithm to obtain the optimal blood glucose concentration prediction model, including: 1) Set the parameters of the fishing optimization algorithm, including: number of fishermen, maximum number of iterations, learning factor and cooperation factor; initialize the position of each fisherman; the position of each fisherman corresponds to a set of parameter values ​​for the blood glucose concentration prediction model; 2) Substitute the current location of each fisherman into the blood glucose concentration prediction model, and input the feature vector of each training sample as the current feature vector into the blood glucose concentration prediction model to obtain the corresponding predicted label; calculate the fitness value of each fisherman's current location based on the true label and predicted label of the training sample: ； In the formula For fitness value, For the first training samples The corresponding predicted label, For the first The true label of each sample; 3) Obtain the optimal fisherman's location and the best fitness value; 4) Based on the optimal individual and the best fitness value, update the position of each fisherman by combining the learning factor and the cooperation factor, and return to step 2) to iterate and optimize until the maximum number of iterations is reached. If so, output the parameter values ​​of the blood glucose concentration prediction model corresponding to the global optimal fisherman position and the global optimal fitness value as the optimal parameter values; substitute the optimal parameter values ​​into the blood glucose concentration prediction model to obtain the optimal blood glucose concentration prediction model.

8. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer programs; The processor is configured to invoke the computer program to perform the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed on an electronic device, causes the electronic device to perform the method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is run on an electronic device, it causes the electronic device to perform the method as described in any one of claims 1 to 7.

Images