54 results about "Stepwise regression" patented technology

The invention relates to a construction method of a metabolic syndrome phlegm syndrome lipid metabolism characteristic diagnosis model. The construction method comprises the following steps: S1, acquiring a serum sample; s2, carrying out lipid metabolite analysis through a full-quantitative lipidomics formula; s3, carrying out primary screening on differential metabolites; s4, optimizing the characteristic indexes by using a random forest algorithm and stepwise regression; and S5, constructing a differential metabolite diagnosis model through Logistic regression analysis. By analyzing differential lipid metabolites of patients with MetS phlegm syndromes and non-phlegm syndromes, the invention reveals that abnormal accumulation of lipid and lipid metabolites may be the core pathological parenchyma of MetS phlegm syndromes. By integrating lipid metabonomics data, using a random forest algorithm, stepwise regression and other methods to screen feature difference lipid metabolites and construct a MetS phlegm syndrome specific diagnosis model, a novel combined biomarker and evidence-based basis are provided for early diagnosis of MetS phlegm syndromes, and a scientific basis can also be provided for objective diagnosis of traditional Chinese medicine phlegm syndromes.

The invention discloses a method for detecting the nitrogen content of organ tissues of different varieties of oilseed rapes based on visible near-infrared spectroscopy, which comprises the following steps: firstly, collecting oilseed rape leaf, shell, stalk and root system samples under the conditions of multiple varieties and multiple nitrogen fertilizer levels, and acquiring spectral data within the range of 430-2500nm by using a visible near-infrared spectroscopy; and a Kjeldahl method is synchronously adopted to measure the real nitrogen content as a reference value. Preprocessing the spectral data, including de-noising, standard normal variable transformation, multivariate scatter correction and derivative processing, so as to weaken the influence of scattering and baseline drift; characteristic wavelengths related to the nitrogen content are screened through stepwise regression, variable projection importance, competitive self-adaptive reweighted sampling, a continuous projection algorithm and other methods, and a random forest model, a support vector machine model, a partial least squares discriminant analysis model, a partial least squares regression model, a support vector regression model, an XGBoost model and other models are combined. And respectively constructing a classification identification model and a regression prediction model.

The invention discloses a method for analyzing influence of climate change and human activity on space-time evolution of water resources. The method comprises the following steps: acquiring meteorological information, hydrological information, land utilization information, soil information, social economic information and water resource information of a to-be-detected area; performing trend analysis on the water resource information to obtain a space-time evolution rule of the water resource information; performing attribution analysis on the time-space evolution rule to obtain influence factors of the time-space evolution of the water resource; obtaining the correlation degree of the influence factors and the space-time evolution rule, and completing the analysis of the climate change and human activity on the space-time evolution of the water resource. According to the method, the influence factors of water resource evolution of the to-be-detected area are researched through the multivariate statistical stepwise regression analysis model, the multicollinearity between the related influence factors of climate change and human activity is reduced through the characteristics of stepwise regression analysis, the calculation accuracy is improved, and the sustainable development of water resources is promoted.

This invention discloses a water treatment dosing control method and system based on quadratic programming, belonging to the field of water treatment process control and optimization technology. It collects historical water treatment operation data and constructs a mechanistic feature set, using the mechanistic feature set as the independent variable and turbidity reduction as the target variable. The turbidity reduction is used to represent the change in flocculation or sedimentation of suspended impurities in the water, and a full-variable regression model is constructed. Variables in the mechanistic feature set are screened, and the full-variable regression model is optimized using a stepwise regression method to obtain a simplified prediction model. Real-time influent water quality parameters are acquired and input into the simplified prediction model. The process of maximizing turbidity reduction in the simplified prediction model is transformed into minimizing a convex loss function, and the convex loss function is iteratively optimized using a hierarchical constrained projection gradient descent method to obtain the optimal dosing scheme. Through mechanism-driven modeling, convex function optimization, and hierarchical constrained projection, intelligent, efficient, and reliable control of the dosing process is achieved.

The present application discloses a wetland vegetation carbon sequestration potential prediction method and system under the influence of future climate change, belongs to the technical field of wetland vegetation carbon sequestration potential prediction, and comprises the following steps: obtaining the net primary productivity of vegetation, the vegetation coverage, the monthly temperature, the precipitation and the atmospheric carbon dioxide concentration in the historical period and the future period; calculating the annual average net primary productivity of vegetation, the coverage and the multi-year average vegetation coverage; calculating the annual average temperature, the precipitation and the carbon dioxide concentration; extracting the wetland vegetation pixel distribution as the research area range; calculating the ratio of the annual average net primary productivity to the vegetation coverage to obtain the wetland vegetation carbon sequestration potential value in the historical period; adopting the multiple stepwise regression analysis method to build a per-pixel wetland vegetation carbon sequestration potential estimation model; and utilizing the annual temperature, the precipitation and the carbon dioxide concentration in the future period to realize the accurate prediction of the wetland vegetation carbon sequestration potential in the research area range in the future period.

The invention belongs to the technical field of medical information processing, and particularly discloses a regression model establishment method and system for detecting related indexes of myocardial cells of Wuzhishan pigs by using blood biochemical indexes, and the method comprises the following steps: collecting blood and heart tissue samples of the Wuzhishan pigs at different month ages, detecting the blood biochemical indexes including liver function, blood fat and blood sugar indexes, and establishing a regression model for detecting the related indexes of myocardial cells of the Wuzhishan pigs. Related indexes of myocardial cells; analyzing the index expression quantity difference, and screening significant difference indexes; then calculating the correlation between the remarkably different blood biochemical indexes and myocardial cell related indexes, and determining high-correlation blood biochemical indexes; and finally, constructing a first version of multiple regression model by respectively taking the myocardial enzyme index and the myocardial cell related index as dependent variables and the high-correlation blood biochemical index as an independent variable, removing outliers and updating the sample set on the basis of the first version of multiple regression model, and constructing a second version of multiple regression model by utilizing the new sample set, and on the basis of the second version of multiple regression model, constructing a third version of multiple regression model by adopting a stepwise regression method. According to the method, related indexes of myocardial cells can be accurately deduced through blood biochemical indexes, and support is provided for Wuzhishan pig heart disease research and model animal research and development.

The invention discloses a method for forecasting anthracnose of wine grapes, and particularly belongs to the technical field of plant disease and insect pest forecasting. The method comprises the following steps: carrying out intra-group correlation analysis on climatic factors of threshold number of days before the wine grape anthracnose attack through SPSS to obtain meteorological factors significantly related to the wine grape anthracnose; through stepwise regression analysis, redundant climate factors in the meteorological factors significantly related to the wine grape anthracnose are removed, a multiple linear regression model is constructed, the meteorological factors significantly related to the wine grape anthracnose are screened out, and multicollinearity is avoided; secondly, inputting meteorological data to be measured into the multiple linear regression model, and outputting a prediction result of the disease index of anthracnose; and finally, forecasting the anthracnose of the wine grapes based on the prediction result of the disease index of the anthracnose. The method can prevent and control anthracnose, improve wine grape quality and reduce economic loss.

This application relates to the field of metal bending component manufacturing technology, and discloses a method for controlling the bending radius in progressive forming of tubular materials by spinning and bending. The method includes the following steps: S1, obtaining an optimal combination of loading parameters affecting the forming quality indicators of the tubular material through orthogonal experiments; S2, establishing a regression function between the bending radius and the spinning and bending deformation parameters based on the optimal combination and the Maximin Latin square test results through stepwise regression analysis; S3, testing the regression function; if the test passes, obtaining the regression model of the bending radius; if the test fails, returning to step S2; S4, controlling the bending radius under the optimal loading parameter conditions using the regression model. This application incorporates a diameter reduction parameter into the control method, significantly improving the accuracy of bending radius control compared to existing geometric control models. It reduces the average error between tubular bending radius prediction and control from 10.28% to 5.2%, offering advantages such as high forming accuracy, wide applicability, and higher reliability.

The invention provides an extreme weather power load prediction method and system, and the method comprises the steps: obtaining historical data, and determining a basic load in extreme weather according to the historical data; constructing a total load decomposition model, stripping initial estimation of a basic load and a random load in the total load according to the total load decomposition model, and obtaining a meteorological load mid-value through multiple regression fitting; according to the meteorological load median, screening core meteorological factors through improved grey correlation analysis, and combining stepwise regression screening and extreme value adaptation correction to obtain an accurate meteorological load; constructing a comprehensive predictive factor set according to the precise meteorological load; and inputting each prediction factor of the comprehensive prediction factor set into the trained improved BP neural network model, and outputting a power load prediction value in extreme weather, thereby effectively improving meteorological load separation precision and extreme weather load prediction precision.

The application discloses a kind of track fusion methods of vehicle, ship: first, the measurement data and real data of target track are preprocessed, the size of data graph is determined by stepwise regression algorithm, then one-dimensional track point data is assembled into two-dimensional track data graph, to achieve the purpose of batch processing measurement data, also can capture the motion correlation between track point data.Second, the interaction between features and the spatial features between data are mined through CIN network structure, then the input of CIN is taken as the input of Informer module, the temporal features between data are mined through Informer module, and finally the fusion result is obtained through LSTM module.Finally, the model is trained and the model file is obtained, and the track fusion task is completed on the validation set.The application can quickly process large track measurement data and accurately calculate the track fusion result, improve the accuracy of track fusion.

The invention discloses a machine learning recommendation method for precision marketing and potential customer mining. The method comprises the following steps: S1, data acquisition: establishing a user tag table based on historical data and acquiring user behavior characteristics; s2, data preprocessing: preprocessing the data, including removing abnormal values, filling missing values, and converting category features into numeric features; s3, automatic box separation: performing variable analysis and automatic box separation on the preprocessed data; s4, feature processing: converting a data table obtained by binning into a numerical value form; s5, dividing a data set: dividing the processed data set into a training set and a test set; s6, stepwise regression is carried out to select features: stepwise logistic regression is used to select model key variables; s7, manual binning and modeling data set preparation are carried out; the method has the characteristics of high automation level and high decision-making efficiency.

The present application relates to the technical field of power system regulation, and more particularly to a power transmission network transient equivalent modeling method and system based on GRNN step-by-step regression, an electronic device and a medium; the method comprises: constructing a power transmission network transient equivalent mathematical model; establishing a detailed simulation model of the power transmission network containing a high proportion of new energy units, and dividing the sample set into a training set and a test set; using the training set to perform step-by-step training on the generalized regression neural network; after the training is completed, calling the regression model, calculating the short-circuit current component of the second branch based on the number of low-penetration units and the voltage correction parameter, and calculating the short-circuit current component of the first branch based on the equivalent impedance, and synthesizing the total short-circuit current. In this way, the technical problem of insufficient short-circuit current calculation accuracy of the existing power transmission network transient equivalent modeling technology in the high proportion of new energy access scene is solved, and the accuracy, adaptability and reliability of the calculation results of the equivalent model are improved.

The invention specifically relates to a significant wave height estimation method based on machine learning. The method comprises the following steps: obtaining a radar echo speed space-time sequence by using a coherent X-band radar; extracting statistical characteristics from the speed space-time sequence; performing Spearman correlation analysis on the statistical characteristic value and the actual significant wave height; screening the characteristic values by using a Spearman correlation coefficient, and constructing a Gaussian process regression and stepwise regression integrated model in combination with significant wave height data; and estimating the significant wave height of the echo to be measured through the integrated model. According to the significant wave height estimation method based on machine learning, key information can be effectively obtained from radar echoes by analyzing the speed space-time sequence of the radar echoes, extracting the statistical characteristics and energy information of the radar echoes and combining with the machine learning technology, and the estimation accuracy of the significant wave height is improved. And a physical correlation model of speed characteristics and wave parameters is constructed by using a machine learning method, so that the accuracy and reliability of inversion are remarkably improved.

The invention discloses a method and equipment for determining a pipeline cleaning strategy, a medium and a product, relates to the technical field of pipeline oil stain cleaning, and aims to solve the problem of how to optimize the pipeline cleaning strategy. The method for determining the pipeline cleaning strategy comprises the following steps: acquiring sample data of pipeline cleaning; based on the sample data, obtaining a cleaning effect multiple regression model, a cleaning effect ridge regression model, a cleaning effect lasso regression model and a cleaning effect stepwise regression model; determining an error evaluation index of the pipeline cleaning strategy determined by each regression model; determining a target cleaning rate model from the cleaning effect multiple regression model, the cleaning effect ridge regression model, the cleaning effect lasso regression model and the cleaning effect stepwise regression model, wherein the target cleaning rate model determines that an error evaluation index of the pipeline cleaning strategy is smaller than a preset index threshold value; and determining a target strategy for pipeline cleaning through the target cleaning rate model.

The invention discloses a blade appearance phenotype multi-dimensional feature-based SPAD value estimation method, system and device, and a medium, mainly relates to the technical field of SPAD value estimation, and is used for solving the problem that the existing scheme introduces too many multi-dimensional features and is easy to cause dimension disaster and overfitting. Comprising the steps of performing foreground and background separation on a digital image to obtain a leaf foreground image; extracting blade appearance phenotype multi-dimensional features from the blade foreground image; through Pearson correlation analysis, screening out significant correlation characteristic parameters of SPAD values from the blade appearance phenotype multi-dimensional characteristics, and screening out an optimal blade appearance phenotype multi-dimensional characteristic combination from the significant correlation characteristic parameters through a stepwise regression method; and obtaining a trained random forest network by using the optimal blade appearance phenotype multi-dimensional feature combination and the SPAD mean value.

This invention provides a method and system for predicting power load under extreme weather conditions. It acquires historical data and determines the base load under extreme weather conditions based on this data. A total load decomposition model is constructed, and based on this model, the base load and random load are initially estimated and separated from the total load. The median meteorological load is obtained through multivariate regression fitting. Based on the median meteorological load, core meteorological factors are screened using improved grey relational analysis, combined with stepwise regression screening and extreme value adaptation correction to obtain the accurate meteorological load. A comprehensive prediction factor set is constructed based on the accurate meteorological load. Each prediction factor in the comprehensive prediction factor set is input into a trained improved BP neural network model, which outputs the predicted power load value under extreme weather conditions, effectively improving the accuracy of meteorological load separation and extreme weather load prediction.

The invention relates to a method for constructing a model for estimating the aboveground biomass of non-uniformly planted rice by using multi-temporal unmanned aerial vehicle image data and an estimation method. The method comprises the following steps: dividing a full life cycle of rice into a vegetative period and a reproductive period by combining a high-resolution unmanned aerial vehicle multispectral image, obtaining a multispectral image of a target area by using an unmanned aerial vehicle in a tillering period, processing the image, extracting a rice pixel proportion, and correcting overground biomass and vegetation indexes by using a rice pixel proportion coefficient to obtain a corrected overground biomass and vegetation index; and then a ternary linear regression model of the vegetation index and the aboveground biomass is established through a multivariate stepwise regression method (SPSS), so that the aboveground biomass estimation precision of the non-uniformly planted rice is remarkably improved.

The invention relates to the technical field of data models, in particular to a yellow rice wine quality prediction method, which comprises the following steps: after establishing a yellow rice wine sensory index evaluation table, respectively carrying out sensory analysis on Fujian yellow rice wine in different years to obtain a comprehensive taste actual score; carrying out correlation and significance analysis on physical and chemical components and sensory indexes of the Fujian yellow wine; performing correlation and significance analysis on the color and sensory indexes of the Fujian yellow wine; performing correlation and significance analysis on volatile matter indexes and sensory indexes of the Fujian yellow wine; selecting parameters with significant levels of less than 0.05 in physical and chemical components, color and volatile matters as independent variables, and establishing a sensory comprehensive evaluation and prediction model by taking the comprehensive taste actual score as a dependent variable and adopting stepwise regression analysis; according to the method, the Fujian yellow rice wine in different years is researched, the quality of the Fujian yellow rice wine is evaluated through sensory analysis, electronic nose analysis and physical and chemical analysis, sensory indexes are quantified, datamation indexes for directly evaluating the quality of the Fujian yellow rice wine are obtained, and management and grading are facilitated.

The application discloses a corn variety screening method suitable for soybean-corn 4:2 and 6:4 strip composite planting. Firstly, newly approved or large-area popularized corn varieties are selected, and different growth environment conditions of edge 1 row, edge 2 row and edge 3 row in each variety plot are created through high-density, non-protection row and relatively wide plot spacing. The agronomic characters, dry matter, yield and constituent factors of the edge 1 row, edge 2 row and edge 3 row are analyzed by using the principal component analysis, fuzzy membership function method, stepwise regression analysis and grey correlation degree analysis method, the key indicators reflecting the edge row advantage of corn are selected and identified, and the edge row advantage of the tested corn varieties is evaluated, so that the corn varieties with strong and weak edge row advantages are screened out, and the corn varieties are recommended as suitable for soybean-corn 4:2 and 6:4 strip composite planting. The screening method is suitable for providing reference for corn varieties of different strip type modes of soybean-corn composite planting and improving the strip composite planting benefit.

The invention discloses a natural gas monthly sales volume main influence factor screening method, system and device, and belongs to the technical field of natural gas data, and the main influence factor screening method comprises the steps: preprocessing natural gas monthly sales volume influence factors and natural gas monthly sales volume data; performing correlation analysis on the relationship between the influence factors and the monthly natural gas sales volume to obtain an influence factor set A; determining contribution values of the influence factors to the natural gas monthly sales prediction result to obtain an influence factor set B; determining importance indexes of the influence factors on the monthly natural gas sales volume to obtain an influence factor set C; determining a minimum public set of the influence factor set A, the influence factor set B and the influence factor set C; and through stepwise regression analysis, screening out main influence factors from the minimum public set. According to the method, the main influence factors of the monthly natural gas sales can be comprehensively and accurately identified and screened, and the precision and interpretability of the prediction model are improved.

The application discloses a prediction model for diagnosing infectious mononucleosis in children and a construction method thereof. The application adopts single factor analysis and logistic stepwise regression to screen independent influencing factors by using confirmed IM inpatient cases and suspected IM but EBV-DNA negative child cases, and adopts multiple factor regression to screen six independent influencing factors of age, ALB, GLB, PLT, LYMPH# and LYMPH% to construct a diagnosis model. Researches show that in the confirmed diagnosis scene, the multi-index combined model with blood routine and liver function has excellent diagnosis efficiency (AUC=0.986), realizes the ideal balance of high sensitivity and high specificity, provides a hierarchical tool for the precise diagnosis of IM in different clinical scenes, and is suitable for the precise diagnosis of inpatient children.

According to the financial scene customer credit evaluation method provided by the invention, an original data set is divided into a test set and a training set according to a preset proportion according to positive and negative sample labels, binning is realized according to a box quantity judgment rule after data preprocessing and feature extraction are performed, and binning adjustment is performed again, so that the customer credit evaluation efficiency is improved. Re-binning processing is carried out according to the feature homogeneity rate and the monotonicity, feature screening is carried out according to the feature homogeneity rate, the IV value and a stepwise regression algorithm after processing is completed, and an optimal credit evaluation score card is calculated and obtained according to a preset weight, tuning parameters and a layered K-fold cross validation method after screening is completed. Compared with the prior art, the method has the advantages that the feature effectiveness and the model generalization ability are improved through technologies such as multi-level feature screening, a layered K-fold cross validation method and dynamic self-adaptive monotonic binning, and the problem of data sparsity and the sensitivity defect of traditional binning to non-monotonic features are solved.

This application relates to the biomedical field and discloses a predictive model and validation method for ART virological failure in HIV / AIDS patients. The model construction method includes the following steps: S1, collecting clinical data from HIV or AIDS patients; S2, performing multiple imputation processing on the clinical data to generate multiple datasets; S3, using a stepwise regression method to screen predictive factors in each dataset; S4, synthesizing the datasets to determine the final predictive factors; S5, establishing a virological failure prediction model based on the final predictive factors. The predictive model constructed by this invention can serve as an effective tool for predicting the risk of ART virological failure in HIV or AIDS patients in clinical practice. The application of this model helps improve treatment effectiveness, optimize resource allocation, and ultimately improve patient treatment outcomes. Furthermore, by continuously collecting new patient data and re-evaluating model performance, the accuracy and practicality of the model can be further improved.

The application discloses a resonator gyroscope temperature compensation method based on stepwise regression and a Gaussian process, and comprises the following steps: screening significant independent variables from preset independent variables through a stepwise regression model; constructing a sample set based on the significant independent variables, training a temperature compensation model; collecting significant variables at a current moment, estimating a zero bias value of the gyroscope through the temperature compensation model; and correcting an output angular velocity of the gyroscope according to the zero bias value. The technical problem of how to effectively suppress the temperature influence on the output zero bias of the hemispherical resonator gyroscope is solved, the zero bias stability of the output result is improved, and the robustness of the temperature compensation model is enhanced.

The application discloses a runoff change attribution analysis method based on underlying surface spatial characteristics, and belongs to the technical field of runoff attribution analysis. First, in the constructed underlying surface spatial characteristic factor set, the key underlying surface spatial characteristic factors capable of effectively explaining the change of Budyko parameters are screened through collinearity test and stepwise regression. Second, an optimal parameter relationship model between the Budyko parameters and the key underlying surface spatial characteristic factors is established. Finally, combined with a water-heat coupling balance equation, the runoff change in different analysis periods is attributed and decomposed, so that the contribution amount of climate change, the contribution amount of underlying surface spatial characteristic change and the absolute contribution proportion are obtained. Through the relationship between the parameters and the underlying surface spatial characteristics, the underlying surface spatial heterogeneity is allowed to participate in the runoff change attribution analysis in a parameterized manner, so that the expression capability of the attribution result to the comprehensive influence of the underlying surface is improved, and guidance is provided for the basin water resource regulation and ecological protection management.

The invention belongs to the technical field of environmental science and chemical engineering, and provides a QSAR (Quantitative Synthetic Aperture Radar) model for predicting adsorption of microplastics to organic pollutants, a construction method and application. A quantitative structure-activity relationship is constructed based on drug physicochemical properties and linear free energy parameters, the ability of polyethylene and polystyrene micro-plastics to adsorb organic pollutants, especially non-steroidal anti-inflammatory drugs, before and after aging is predicted, and the ability of the polyethylene and polystyrene micro-plastics to adsorb organic pollutants, especially non-steroidal anti-inflammatory drugs, is predicted by taking a distribution coefficient (Kd) of the organic pollutants in the micro-plastics as a dependent variable and taking the physicochemical property parameters of the organic pollutants as independent variables. A retreating stepwise regression method is adopted to screen variables and determine model parameters, on the premise of ensuring model interpretability, an optimal independent variable combination is screened out, the internal relation between the variables is accurately recognized, redundancy with high correlation is effectively eliminated, the collinearity degree between the variables is remarkably reduced, and the method is suitable for large-scale popularization and application. The obtained multiple QSAR model formulas are used for predicting organic pollutants adsorbed by micro-plastics, especially non-steroidal anti-inflammatory drugs, and the prediction accuracy and precision are improved.

The invention discloses a stepwise regression analysis method-based college student physique test score prediction model, and belongs to the technical field of sports and metering economics. The method comprises the following steps: collecting and preprocessing college student physique test multi-dimensional index data; constructing a multiple linear regression model between the total physical test score and a plurality of explanatory variables; screening significant variables by adopting a stepwise regression analysis method, and optimizing a model structure; performing multi-collinearity, heteroscedasticity and sequence correlation test and correction on the model; and finally verifying the validity of the variable set by using a random forest model. According to the method, key influence factors can be automatically identified from numerous physical indexes, a prediction model with high goodness of fit and high interpretation is constructed, accurate prediction of student physical scores is realized, and a scientific basis is provided for physical health management of colleges and universities.

The application provides a target detection method and device based on stepwise regression, which comprises the following steps: obtaining a picture to be detected, wherein the picture to be detected comprises a target picture and a target frame corresponding to the target picture, the target frame comprises a first target frame and a second target frame with a pixel number greater than the first target frame; performing scale transformation on the first target frame according to a preset ratio to obtain a third target frame with the scale size of the second target frame; obtaining a candidate target frame according to the third target frame and the second target frame, and inputting the picture to be detected and the candidate target frame into a specific target detection model to obtain a target prediction result output by the specific target detection model. The first target frame is subjected to scale transformation according to the scale size of the second target frame to enlarge the feature expression capability of the first target; then the specific target detection model is used to detect the enlarged first target, and a regression mode from large to small is adopted for the target frame, so that the precision of small target detection is greatly improved.

The invention provides an earthquake frequency prediction method based on delay correlation, and belongs to the field of earthquake frequency prediction, and the method comprises the steps: obtaining earthquake catalog data in a fracture system region, generating a complete earthquake catalog, carrying out the statistics of an earthquake frequency time sequence based on a preset earthquake magnitude range and a time unit, and calculating the delay cross correlation coefficient of the earthquake frequency time sequence, selecting a maximum delay cross correlation coefficient and a corresponding lag order from the seismic frequency time sequence according to the delay cross correlation coefficients; and according to the seismic frequency time sequence, the maximum delay cross correlation coefficient and the corresponding lagging order, constructing a seismic frequency time sequence delay correlation influence path network, through stepwise regression, obtaining a regression model, and screening out a regression model with the lagging order less than 0 to carry out seismic frequency prediction. According to the method, the problem that high-precision and quantifiable prediction cannot be carried out on the seismic activity frequency in a fracture system due to the lack of quantitative statistical prediction based on fracture seismic delay correlation in the prior art is solved.