A method and device for mining association patterns of supply and demand resources of a ship

Abstract

The application discloses a ship supply-demand resource association mode mining method and device, and the method comprises the following steps: collecting a ship resource historical data set; the ship resource historical data set comprises a ship-carrying resource data set and a ship demand data set; the ship-carrying resource data set is a technical state data set of equipment resources carried by a ship when the ship sets sail; the ship resource historical data set is preprocessed to obtain a pretreated data set; the pretreated data set is subjected to association mode mining processing to obtain a ship resource demand prediction model.

Description

Technical Field

[0001] This invention relates to the fields of industrial data processing, strategy optimization processing, and big data mining, specifically to a method and apparatus for mining correlation patterns of ship supply and demand resources. Background Technology

[0002] With the rapid development of the global shipping industry, the availability of equipment resources and the matching of needs during ship navigation have become key factors affecting shipping efficiency and safety. Currently, ship resource management mainly relies on manual experience or simple rule matching, which presents the following technical problems:

[0003] Insufficient accuracy in resource demand forecasting: There is a complex nonlinear relationship between the technical status of shipboard equipment (such as operating parameters, wear and tear) and actual demand. Traditional methods are unable to capture the dynamic impact of multi-dimensional technical attributes, resulting in large deviations in demand forecasting.

[0004] The reliance on a single clustering method leads to a one-sided approach to pattern mining: Existing technologies often use a single clustering algorithm to analyze ship resource data, ignoring the differences in the adaptability of different algorithms to sequential data and graph structure data, making it difficult to fully mine potential patterns in technical status data.

[0005] Limited adaptability of correlation models: For the correlation analysis between technical requirements and carrying resources, traditional linear regression models cannot handle the collinearity problem of high-dimensional variables, while single nonlinear models are difficult to balance prediction accuracy and generalization ability, resulting in limited applicability of models in complex ship scenarios.

[0006] The data preprocessing stage is crude: ship data often contains noise and redundant information, and existing technologies lack targeted cleaning and standardization processes, which directly affects the reliability and efficiency of subsequent models. Summary of the Invention

[0007] This invention primarily addresses the problem that existing rules for matching equipment resources with demand during ship navigation are simplistic and lack accuracy in supply and demand forecasting. This invention discloses a method and apparatus for mining correlation patterns of ship supply and demand resources.

[0008] In a first aspect, this invention discloses a method for mining correlation patterns of ship supply and demand resources, comprising:

[0009] S1, Collect historical data set of ship resources; the historical data set of ship resources includes a data set of ship-carried resources and a data set of ship requirements; the data set of ship-carried resources is a data set of the technical status of equipment resources carried by ships when they go to sea; the data set of technical status includes discrete sequence data of the status of various technical attributes; the data set of ship requirements is a data set of technical requirements for equipment resources needed by ships; the data set of technical requirements includes discrete sequence data of various technical attributes required by ships.

[0010] S2, preprocess the historical data set of ship resources to obtain a preprocessed data set;

[0011] S3, perform association pattern mining on the preprocessed dataset to obtain a ship resource demand prediction model; the ship resource demand prediction model is used to predict ship demand data based on the received real-time resource dataset carried by ships.

[0012] The process of preprocessing the historical data set of ship resources to obtain a preprocessed dataset includes:

[0013] S21, perform data cleaning on the historical data set of ship resources to obtain the first data set;

[0014] S22, perform data reduction processing on the first dataset to obtain the second dataset;

[0015] S23, perform pattern reduction on the second data to obtain a preprocessed dataset.

[0016] The process of performing association pattern mining on the preprocessed dataset to obtain a ship resource demand prediction model includes:

[0017] S31, perform fusion and clustering processing on the technical status dataset in the preprocessed dataset to obtain a set of cluster centers;

[0018] S32, perform demand modeling processing on the cluster center set and the technology demand dataset in the preprocessed dataset to obtain the ship resource demand prediction model.

[0019] The step of fusing and clustering the technical status dataset in the preprocessed dataset to obtain a set of cluster centers includes:

[0020] S311, A sequence-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain a first set of cluster centers and a corresponding set of the number of category elements; the set of the number of category elements includes the total number of elements contained in the cluster category corresponding to each first cluster center in the first set of cluster centers; the first set of cluster centers includes several cluster centers.

[0021] S312, A graph-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain a second set of cluster centers and a corresponding set of the number of category elements;

[0022] S313, A clustering algorithm based on a probability model is used to cluster the technical status dataset in the preprocessed dataset to obtain the third cluster center set and the corresponding set of the number of category elements;

[0023] S314, A suffix tree-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain the fourth cluster center set and the corresponding set of the number of category elements;

[0024] S315, perform a fusion calculation on the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set to obtain a fusion center set.

[0025] The process of merging the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set to obtain a merged center set includes:

[0026] S3151, the cluster centers of the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set are sorted from high to low according to the total number of elements, respectively, to obtain a corresponding sequence for each cluster center set; the sequence includes several cluster centers;

[0027] S3152, perform fusion calculation on the corresponding sequences of all cluster center sets to obtain the fusion center set.

[0028] The process of performing demand modeling on the cluster center set and the technology demand dataset in the preprocessed dataset to obtain a ship resource demand prediction model includes:

[0029] Using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, a multivariate linear regression analysis is performed on the multivariate independent variable and the multivariate dependent variable to obtain the regression analysis results and regression model for each dependent variable.

[0030] For the dependent variable whose regression analysis result is yes, determine the regression model of the dependent variable, which is the ship resource demand prediction model of the dependent variable;

[0031] For the dependent variable whose regression analysis result is negative, perform multivariate correlation analysis to obtain the correlation model of the corresponding dependent variable;

[0032] The correlation model of the dependent variable is determined to be the ship resource demand prediction model of the dependent variable.

[0033] The step involves using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, performing multivariate linear regression analysis on the multivariate independent and dependent variables to obtain the regression analysis results and regression model for each dependent variable, including:

[0034] Using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, regression analysis is performed on each of the multivariate dependent variables with the multivariate independent variables to obtain the corresponding regression model.

[0035] Using the regression model for each dependent variable, the cluster center set is calculated and processed to obtain the corresponding set of regression values;

[0036] An F-test is performed on the set of regression values ​​and the corresponding values ​​of the dependent variable in the technical requirements dataset of the preprocessed dataset to obtain the test results;

[0037] If the test result is passed, the regression analysis result of the dependent variable is determined to be yes; if the test result is failed, the regression analysis result of the dependent variable is determined to be no.

[0038] A second aspect of the present invention discloses a device for mining correlation patterns of ship supply and demand resources, the device comprising:

[0039] Memory containing executable program code;

[0040] A processor coupled to the memory;

[0041] The processor calls the executable program code stored in the memory to execute the method for mining the correlation pattern of ship supply and demand resources.

[0042] In a third aspect of this invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are invoked by a computer, they are used to execute the method for mining the correlation pattern of ship supply and demand resources.

[0043] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the method for mining the correlation patterns of ship supply and demand resources.

[0044] The beneficial effects of this invention are as follows:

[0045] This invention integrates multi-source clustering algorithms and multivariate regression analysis to fully explore the potential correlation between ship-borne resources and demand, making the prediction results more consistent with actual navigation scenarios and providing accurate basis for ship resource allocation.

[0046] This invention employs four clustering algorithms based on sequence, graph, probabilistic models, and suffix trees, combining the advantages of various algorithms through fusion computing to overcome the limitations of single algorithms and more comprehensively capture the deep patterns in technical status data. For different dependent variables, linear regression and multivariate association analysis (fusion of ridge regression, Lasso regression, etc.) are used respectively, preserving the interpretability of linear relationships while enhancing predictive capabilities in complex scenarios through nonlinear fusion, balancing accuracy and generalization. Through a three-level preprocessing workflow of data cleaning, reduction, and pattern reduction, noise is effectively removed and data dimensionality is compressed, providing a high-quality data foundation for subsequent analysis and reducing model computational complexity. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention. Detailed Implementation

[0048] To better understand the content of this invention, an embodiment is provided here.

[0049] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention.

[0050] In a first aspect, this invention discloses a method for mining correlation patterns of ship supply and demand resources, comprising:

[0051] S1, Collect historical data set of ship resources; the historical data set of ship resources includes a data set of ship-carried resources and a data set of ship requirements; the data set of ship-carried resources is a data set of the technical status of equipment resources carried by ships when they go to sea; the data set of technical status includes discrete sequence data of the status of various technical attributes; the data set of ship requirements is a data set of technical requirements for equipment resources needed by ships; the data set of technical requirements includes discrete sequence data of various technical attributes required by ships.

[0052] S2, preprocess the historical data set of ship resources to obtain a preprocessed data set;

[0053] S3, perform association pattern mining on the preprocessed dataset to obtain a ship resource demand prediction model; the ship resource demand prediction model is used to predict ship demand data based on the received real-time resource dataset carried by ships.

[0054] The process of preprocessing the historical data set of ship resources to obtain a preprocessed dataset includes:

[0055] S21, perform data cleaning on the historical data set of ship resources to obtain the first data set;

[0056] S22, perform data reduction processing on the first dataset to obtain the second dataset;

[0057] S23, perform pattern reduction on the second data to obtain a preprocessed dataset.

[0058] The data cleaning process includes outlier removal and smoothing filtering.

[0059] The data reduction process can be implemented using feature reduction and sample reduction.

[0060] The mode reduction process includes:

[0061] S231, for the technical status dataset and technical demand dataset, an approximation matrix is ​​constructed using the data sequence of each type of dataset in the dataset as the row vector; a collection matrix is ​​constructed using the data collection information of each type of data sequence in the dataset; the row vector of the collection matrix is ​​the data collection information of a type of data sequence; a pattern matching model is constructed using the approximation matrix as the dependent variable and the collection matrix as the independent variable.

[0062] S232, Solve the pattern matching model to obtain the matching model for each type of dataset; the expression of the matching model is f(x) = xA, where x is the acquisition matrix of the input of the matching model, and matrix A is obtained by solving the pattern matching model;

[0063] S233: Using the matching model for each type of dataset, the data collection information of the data sequence of the corresponding type of dataset is calculated and processed to obtain the solution matrix;

[0064] S234, For each data point in the solution matrix, subtract it from the corresponding data in the approximation matrix that shares the same row and column as the data in the solution matrix, and calculate the absolute value to obtain the difference value of the data;

[0065] S235, delete the data whose difference value is greater than the set difference threshold from the second dataset;

[0066] S236, Perform S231 to S235 on the technical status dataset and technical requirement dataset respectively to obtain a preprocessed dataset.

[0067] The process of performing association pattern mining on the preprocessed dataset to obtain a ship resource demand prediction model includes:

[0068] S31, perform fusion and clustering processing on the technical status dataset in the preprocessed dataset to obtain a set of cluster centers;

[0069] S32, perform demand modeling processing on the cluster center set and the technology demand dataset in the preprocessed dataset to obtain a ship resource demand prediction model;

[0070] The step of fusing and clustering the technical status dataset in the preprocessed dataset to obtain a set of cluster centers includes:

[0071] S311, A sequence-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain a first set of cluster centers and a corresponding set of the number of category elements; the set of the number of category elements includes the total number of elements contained in the cluster category corresponding to each first cluster center in the first set of cluster centers.

[0072] For the subsequent second to fourth clustering algorithms, the definition of the set of number of elements in each category is the same as the definition above;

[0073] S312, A graph-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain a second set of cluster centers and a corresponding set of the number of category elements;

[0074] S313, A clustering algorithm based on a probability model is used to cluster the technical status dataset in the preprocessed dataset to obtain the third cluster center set and the corresponding set of the number of category elements;

[0075] S314, A suffix tree-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain the fourth cluster center set and the corresponding set of the number of category elements;

[0076] S315, perform a fusion calculation on the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set to obtain a fusion center set;

[0077] The cluster center set is the set of center points for each category obtained by the clustering algorithm, and the center point of each category is the cluster center.

[0078] The second to fourth cluster center sets each include several cluster centers;

[0079] The process of merging the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set to obtain a merged center set includes:

[0080] S3151, the cluster centers of the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set are sorted from high to low according to the total number of elements, so as to obtain the corresponding sequence of each cluster center set;

[0081] S3152, perform fusion calculation on the corresponding sequences of all cluster center sets to obtain the fusion center set.

[0082] The expression for the fusion calculation is:

[0083]

[0084] Where, μ ijk Let μ be the k-th element of the j-th cluster center in the sequence corresponding to the i-th cluster center set, where N represents the total number of cluster centers in the sequence participating in the calculation. 0j β is the mean of the k-th element of the j-th cluster center in the corresponding sequence of all cluster center sets. jk ω1 and ω2 are preset weighting factors, which can take values ​​of 0.3 and 0.6 respectively.

[0085] The first part of the fusion calculation expression fuses the normalized features of each cluster center element through multiplication and logarithmic operations. This preserves the relative distribution relationship of the outputs of different clustering algorithms and eliminates dimensional differences through geometric normalization, preventing a single algorithm from dominating the results due to numerical scale bias. The second part strengthens the suppression of abnormal cluster centers that deviate from the overall trend by multiplying the mean deviation term and the logarithmic weight, thus improving the stability of the fusion result. Two weighting factors can be used to adjust the weights of "distribution consistency" and "bias tolerance" according to the actual data characteristics. When the effective patterns in the data are clear, ω1 can be increased to highlight consensus features; when there are algorithmic disturbances or noise, ω2 can be increased to enhance anti-interference capabilities and flexibly adapt to different ship data scenarios. By combining the geometric mean (Nth root) and the arithmetic mean (cumulative summation), the excessive influence of extreme values ​​on the overall fusion result is avoided, and the complementary information of multiple algorithms is fully utilized. This allows the final fusion center to more comprehensively reflect the deep patterns of ship resource technical status data, providing more reliable basic features for subsequent demand forecasting.

[0086] The process of performing demand modeling on the cluster center set and the technology demand dataset in the preprocessed dataset to obtain a ship resource demand prediction model includes:

[0087] Using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, a multivariate linear regression analysis is performed on the multivariate independent variable and the multivariate dependent variable to obtain the regression analysis results and regression model for each dependent variable.

[0088] For the dependent variable whose regression analysis result is yes, determine the regression model of the dependent variable, which is the ship resource demand prediction model of the dependent variable;

[0089] For the dependent variable whose regression analysis result is negative, perform multivariate correlation analysis to obtain the correlation model of the corresponding dependent variable;

[0090] The correlation model of the dependent variable is determined to be the ship resource demand prediction model of the dependent variable.

[0091] The step involves using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, performing multivariate linear regression analysis on the multivariate independent and dependent variables to obtain the regression analysis results and regression model for each dependent variable, including:

[0092] Using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, regression analysis is performed on each of the multivariate dependent variables with the multivariate independent variables to obtain the corresponding regression model.

[0093] Using the regression model for each dependent variable, the cluster center set is calculated and processed to obtain the corresponding set of regression values;

[0094] An F-test is performed on the set of regression values ​​and the corresponding values ​​of the dependent variable in the technical requirements dataset of the preprocessed dataset to obtain the test results;

[0095] If the test result is passed, the regression analysis result of the dependent variable is determined to be yes; if the test result is failed, the regression analysis result of the dependent variable is determined to be no.

[0096] The multivariate association analysis process includes:

[0097] A fusion regression analysis is performed on the dependent variable and multiple independent variables to obtain the corresponding set of regression functions; the fusion regression analysis includes ridge regression analysis, Lasso regression analysis, and least squares regression analysis; the set of regression polynomials includes the regression function obtained by each regression method and the corresponding performance index; the performance index includes the coefficient of determination and the mean squared error.

[0098] The regression function set is subjected to a first fusion calculation to obtain the correlation model corresponding to the dependent variable;

[0099] The expression for the first fusion calculation is:

[0100]

[0101] Where x is a multivariate independent variable, r(x) is the association model, and H i Let y represent the i-th order Hermitian function. i and e i Let f represent the coefficient of determination and mean squared error of the regression function obtained by the i-th regression method, respectively. i Let represent the regression function obtained by the i-th regression method. The first to third regression methods are ridge regression analysis, Lasso regression analysis, and least squares regression analysis, respectively.

[0102] The first fusion calculation, through The ratio of the coefficient of determination to the mean squared error quantifies the overall performance of each regression method, with the method exhibiting superior performance (high coefficient of determination, low error) receiving higher weight in the fusion. Simultaneously, by leveraging the Hermitian function's orthogonal decomposition capability of the feature space, the advantages of each regression method in different feature dimensions can be selectively extracted (e.g., anti-collinearity of ridge regression, feature selection by Lasso, and unbiasedness of least squares), avoiding the limitations of a single method. The fusion of the three regression methods preserves the interpretability of the linear model (through least squares and ridge regression) while achieving feature dimensionality reduction through Lasso, reducing interference from redundant variables. For high-dimensional, multicollinear data in ship resource demand forecasting (e.g., the correlation of multiple equipment technical indicators), this expression maintains the model's traceability to key influencing factors while ensuring prediction accuracy. As an orthogonal polynomial, the Hermitian function can capture different degrees of nonlinearity through order differences. Combined with the regression function, it can effectively fit the complex mapping relationship between ship resource operating status and demand (e.g., the nonlinear growth relationship between equipment wear rate and spare parts demand), improving the model's adaptability to dynamic navigation scenarios.

[0103] The regression model is a multinomial model; the multiple linear regression analysis can be performed using linear regression analysis, specifically, the least squares method can be used to solve for the coefficients of the regression equation.

[0104] The calculation and processing of the F-test statistic shall be carried out in accordance with the calculation formula of the F-test statistic.

[0105] All variables in all formulas of this invention have been dimensionless before calculation.

[0106] The log operation in this invention is a base-2 logarithmic operation.

[0107] A second aspect of the present invention discloses a device for mining correlation patterns of ship supply and demand resources, the device comprising:

[0108] Memory containing executable program code;

[0109] A processor coupled to the memory;

[0110] The processor calls the executable program code stored in the memory to execute the method for mining the correlation pattern of ship supply and demand resources.

[0111] In a third aspect of this invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are invoked by a computer, they are used to execute the method for mining the correlation pattern of ship supply and demand resources.

[0112] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the method for mining the correlation patterns of ship supply and demand resources.

[0113] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for mining correlation patterns of ship supply and demand resources, characterized in that, include: S1, Collect historical data set of ship resources; the historical data set of ship resources includes ship carrying resource dataset and ship demand dataset; The ship-carried resource dataset is a dataset of the technical status of the equipment resources carried by the ship when it goes to sea; the technical status dataset includes discrete sequence data of the status of various technical attributes; the ship demand dataset is a dataset of the technical requirements of the equipment resources required by the ship; the technical demand dataset includes discrete sequence data of various technical attributes required by the ship. S2, preprocess the historical data set of ship resources to obtain a preprocessed data set; S3, perform association pattern mining on the preprocessed dataset to obtain a ship resource demand prediction model.

2. The method for mining correlation patterns of ship supply and demand resources as described in claim 1, characterized in that, The process of preprocessing the historical data set of ship resources to obtain a preprocessed dataset includes: S21, perform data cleaning on the historical data set of ship resources to obtain the first data set; S22, perform data reduction processing on the first dataset to obtain the second dataset; S23, perform pattern reduction on the second data to obtain a preprocessed dataset.

3. The method for mining correlation patterns of ship supply and demand resources as described in claim 2, characterized in that, The process of performing association pattern mining on the preprocessed dataset to obtain a ship resource demand prediction model includes: S31, perform fusion and clustering processing on the technical status dataset in the preprocessed dataset to obtain a set of cluster centers; S32, perform demand modeling processing on the cluster center set and the technology demand dataset in the preprocessed dataset to obtain the ship resource demand prediction model.

4. The method for mining correlation patterns of ship supply and demand resources as described in claim 3, characterized in that, The step of fusing and clustering the technical status dataset in the preprocessed dataset to obtain a set of cluster centers includes: S311, A sequence-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain a first set of cluster centers and a corresponding set of the number of category elements; the set of the number of category elements includes the total number of elements contained in the cluster category corresponding to each first cluster center in the first set of cluster centers; the first set of cluster centers includes several cluster centers. S312, A graph-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain a second set of cluster centers and a corresponding set of the number of category elements; S313, A clustering algorithm based on a probability model is used to cluster the technical status dataset in the preprocessed dataset to obtain the third cluster center set and the corresponding set of the number of category elements; S314, A suffix tree-based clustering algorithm is used to cluster the technical status dataset in the preprocessed dataset to obtain the fourth cluster center set and the corresponding set of the number of category elements; S315, perform a fusion calculation on the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set to obtain a fusion center set.

5. The method for mining correlation patterns of ship supply and demand resources as described in claim 4, characterized in that, The process of merging the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set to obtain a merged center set includes: S3151, the cluster centers of the first cluster center set, the second cluster center set, the third cluster center set, and the fourth cluster center set are sorted from high to low according to the total number of elements, respectively, to obtain a corresponding sequence for each cluster center set; the sequence includes several cluster centers; S3152, perform fusion calculation on the corresponding sequences of all cluster center sets to obtain the fusion center set.

6. The method for mining correlation patterns of ship supply and demand resources as described in claim 5, characterized in that, The process of performing demand modeling on the cluster center set and the technology demand dataset in the preprocessed dataset to obtain a ship resource demand prediction model includes: Using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, a multivariate linear regression analysis is performed on the multivariate independent variable and the multivariate dependent variable to obtain the regression analysis results and regression model for each dependent variable. For the dependent variable whose regression analysis result is yes, determine the regression model of the dependent variable, which is the ship resource demand prediction model of the dependent variable; For the dependent variable whose regression analysis result is negative, perform multivariate correlation analysis to obtain the correlation model of the corresponding dependent variable; The correlation model of the dependent variable is determined to be the ship resource demand prediction model of the dependent variable.

7. The method for mining correlation patterns of ship supply and demand resources as described in claim 6, characterized in that, The step involves using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, performing multivariate linear regression analysis on the multivariate independent and dependent variables to obtain the regression analysis results and regression model for each dependent variable, including: Using the cluster center set as the multivariate independent variable and the technology demand dataset in the preprocessed dataset as the multivariate dependent variable, regression analysis is performed on each of the multivariate dependent variables with the multivariate independent variables to obtain the corresponding regression model. Using the regression model for each dependent variable, the cluster center set is calculated and processed to obtain the corresponding set of regression values; An F-test is performed on the set of regression values ​​and the corresponding values ​​of the dependent variable in the technical requirements dataset of the preprocessed dataset to obtain the test results; If the test result is passed, the regression analysis result of the dependent variable is determined to be yes; if the test result is failed, the regression analysis result of the dependent variable is determined to be no.

8. A device for mining correlation patterns of ship supply and demand resources, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the method for mining the correlation pattern of ship supply and demand resources as described in any one of claims 1 to 7.

9. A computer-storable medium, characterized in that, The computer storage medium stores computer instructions, which, when invoked by the computer, are used to execute the method for mining the correlation patterns of ship supply and demand resources as described in any one of claims 1 to 7.

10. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the method for mining the correlation patterns of ship supply and demand resources as described in any one of claims 1 to 7.

Images