28 results about "Variable screening" patented technology

The invention provides an interpretable machine learning-based giant coronary artery tumor prediction method in Kawasaki disease, and belongs to the technical field of biological information processing. Comprising the following steps: S1, collecting hospitalized KD child patient cases by adopting a retrospective queue research method, and excluding cases which do not meet requirements; s2, coronary artery lesion evaluation and classification; s3, performing variable preprocessing; s4, constructing a machine learning model; s5, performing variable screening; s6, model interpretation; s7, determining an optimal intervention threshold value; and S8, developing an online prediction tool. According to the method, an SVM kernel lab model with the optimal performance is screened out through multi-model comparison, and a group of key prediction molecules are determined in combination with an SHAP method and recursive feature elimination; based on a multi-center data verification model, the model applicability is improved; and early recognition and layered management of high-risk child patients can be realized. A webpage tool is developed, and a patient with high risk is suggested to intervene in advance.

The invention provides a heat supply system thermal load prediction method based on time sequence compensation and hybrid learning, and the method comprises the steps: constructing a thermal load prediction input variable screening mechanism based on two-dimensional analysis, and achieving the scientific screening of input variables through the dual verification of a Pearson's correlation coefficient matrix and a significance test; a thermal load-temperature time sequence synchronization system based on dynamic time shift compensation is innovatively designed, and the problem of time sequence dislocation caused by instability of manual experience adjustment in a central heating system is solved; an ES-LSTM collaborative prediction architecture is provided, and a three-level prediction model of'trend decomposition-random learning-dynamic weighting 'is established. According to the method, statistics significance test and thermodynamic mechanism analysis are combined in a breakthrough manner, a dynamic compensation system with a time sequence self-correction capability is innovatively researched and developed, a hybrid model collaborative prediction mechanism is constructed, and a complete technical chain from data preprocessing to prediction model architecture is formed.

The invention discloses a fault diagnosis system and method based on a production process causal graph model, and the system comprises a data collection and preprocessing module which is used for collecting and processing fault feature data, fault result data and auxiliary data in the operation process of a production process; the agent variable screening module is used for screening fault agent variables by calculating the correlation between auxiliary data and unobserved hybrid variables in combination with domain knowledge; the causal effect estimation module obtains a causal effect by adopting a mode of combining a tool variable method and deep dual machine learning; the causal graph model is constructed according to a causal effect estimation result and a Bayesian network to represent a causal relationship and a conditional dependency relationship among a fault feature variable, a fault result variable and a fault agent variable; the fault diagnosis module; inputting the operation data collected in real time into the constructed causal graph model, and evaluating the influence degree of different factors on the fault diagnosis result; the method effectively overcomes the interference of hybrid variables in the production process of the manufacturing industry, and improves the fault diagnosis precision and reliability.

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 invention relates to the technical field of thickeners, and provides a thickener feeding amount prediction method and device, equipment and a medium, and the method comprises the steps: obtaining key parameter data of an upstream production process of a thickener; performing preprocessing, performing variable screening to determine a target variable, and calculating the maximum delay time between the target variable and the feeding amount of the thickener to obtain preprocessed data; constructing a dual-path feature extraction network for extracting periodic features and trend features; performing characteristic decomposition on the preprocessed data to obtain an initial trend characteristic and a periodic characteristic; inputting into a dual-path feature extraction network to output a target trend feature and a target period feature; performing model training to obtain a target thickener feeding amount prediction model; and on the basis of a multi-step prediction strategy, predicting the feeding amount data of the to-be-predicted thickener in the future preset duration through the target thickener feeding amount prediction model. According to the technical scheme, the change trend of the feeding amount of the thickener is effectively monitored, and the prediction accuracy is improved.

The invention discloses a time-resolved metal material tensile strength detection device and method, and belongs to the technical field of metal material mechanical property detection. The method comprises the following steps: 1, configuring sample spectrum acquisition by using a time-resolved metal material tensile strength detection device; 2, respectively constructing a spectrum matrix Mn under different delay time; 3, sequentially carrying out saturation removal, abnormity removal, background removal and normalization on the spectrum matrix to form a sample initial feature vector; 4, respectively acquiring training set feature matrixes under different delay times, and forming training set feature input of the prediction model by adopting a variable screening and recombination mode; 5, performing multi-fold cross validation by using the randomly selected part of samples to adjust hyper-parameters so as to obtain a prediction model; the trained prediction model is used for metal material tensile strength detection; compared with the prior art, the technical problem of low accuracy in metal material tensile strength detection is solved by adopting the time-resolved spectrum.

The invention discloses a prognostic scoring method and system for predicting advanced biliary tract malignant tumor immunotherapy, and belongs to the technical field of advanced biliary tract malignant tumor immunotherapy. The method comprises the following steps: firstly, generating random replacement copies for real variables to obtain shadow variables, and classifying the real variables into related classes by comparing importance scores of the random replacement copies and the shadow variables; real variables in related classes are screened based on LASSO regression coefficients, and a variable set is formed; combining the variables in the set, and screening out a variable combination corresponding to a minimum AIC value model according to an AIC criterion to obtain a prognosis key variable set; calculating an index score according to the COX regression coefficient of each real variable; and finally, adding the index scores of the variables in the prognosis key variable set of each patient to obtain a prognosis BICCAPS score. Through three-step progressive screening of the random forest algorithm, the LASSO and the optimal subset method, the problems of variable screening, colinearity and model conciseness in high-dimensional clinical data modeling are solved.

The application discloses a kind of threshold beam profile structure optimization system and method, the system includes initial modeling module, key variable screening module and stage optimization module.Initial modeling module is used to fill the stiffener model in the cavity of the base profile section, generates initial section and finite element model.Key variable screening module passes through calculating the sensitivity of section shape variable to vehicle body bending modal frequency and side column collision intrusion, and screens out key design variable.Stage optimization module is based on key design variable, and section optimization, length optimization and stiffener optimization are sequentially executed, and the optimized stiffener layout is obtained.The application solves the problem that traditional method is difficult to cooperatively optimize section shape, length and stiffener parameters to balance lightweight and multiple performance requirements, can significantly reduce the weight of threshold beam, and shorten the optimization cycle.

The invention discloses an intelligent galvanized pipe corrosion resistance prediction method and system. The method comprises the steps of data acquisition, galvanized pipe corrosion resistance prediction model design, parameter adjustment and galvanized pipe corrosion resistance prediction. The invention belongs to the field of data processing, and particularly relates to an intelligent galvanized pipe corrosion resistance prediction method and system.According to the scheme, a time accumulation effect item is introduced to design a mixed structure different line model, the long-term corrosion dynamic state is accurately described, and the corrosion resistance factor influence capture precision is improved; performing synchronous variable screening by designing a target function; the sectional corrosion loss function uses a power potential reduction tail section for the distortion amount to reduce the interference of the distortion amount; constructing a performance-by-performance function, quantifying the adaptability of corrosion data, and avoiding distortion interference; combining a goodness-of-fit item and a complexity penalty item to realize penalty parameter tuning; through learning rate adjustment based on loss camber, conventional corrosion and distortion amount areas are accurately adapted, and then the corrosion resistance prediction effect is improved.

The application discloses a variable importance screening method based on multi-fidelity sensitivity error, and belongs to the technical field of adapter variable screening. S1, M variables to be analyzed are acquired; S2, a high-fidelity sample set and a low-fidelity sample set are respectively collected as a training set; S3, a multi-fidelity proxy model is constructed, and a first importance measure result of the M variables to be analyzed is calculated; S4, a contribution value of the high-fidelity sample set to the first importance measure result is calculated; S5, a new sample point is solved; S6, error improvement amplitudes of the new sample point to a high-fidelity sensitivity index and a low-fidelity sensitivity index are respectively calculated, and a fidelity type of the new sample point is obtained; S7, the multi-fidelity proxy model is updated, and a second importance measure result of the M variables to be analyzed is calculated; S8, in a case that the second importance measure result converges, the second importance measure result is taken as a target importance measure result, and in a case that the second importance measure result does not converge, the step S4 is jumped. The efficiency of adapter variable importance screening is improved.

The present invention relates to the technical field of key parameter prediction in petrochemical production processes, and more specifically, to a method for modeling key indicators of catalytic cracking units that integrates time series feature extraction. The method comprises: step S1, obtaining catalytic unit production data for preprocessing; step S2, screening input feature variables using a random forest method to generate sample data; step S3, dividing the sample data into different operating modes based on a principal component analysis method to form a plurality of corresponding sub-mode samples; step S4, integrating relative position encoding to obtain encoded input information under each sub-mode; step S5, constructing a multi-layer encoding module based on an attention mechanism and a feedforward neural network, and transforming the output of the multi-layer encoding module into product yield data; step S6, adjusting some model structure hyperparameters to generate a prediction model, predict the product yield, and process the output data. The catalytic unit data model generated by the present invention can improve the prediction accuracy and stability of product yield.

The invention relates to a wheat nitrogen content detection method based on SG-CWT coupled SPA feature dynamic dimension reduction, which comprises the following steps: constructing a dynamic feature purification mechanism, and carrying out efficient feature compression and significant variable screening; a dynamic modeling and robust optimization mechanism is introduced to enhance the adaptability to the time sequence and variety difference; a lightweight modeling framework is provided to reduce model calculation overhead and operation delay. According to the method, a signal enhancement-feature purification-intelligent modeling cascade optimization system is constructed, multi-scale feature deep mining is achieved, compared with a traditional method, the number of feature wavebands is increased by 83%, a dynamic dimension reduction mechanism under the precision loss constraint is provided, the RMSE amplification smaller than 5% is used as a threshold value to screen the optimal waveband combination, 97.2% of model performance is reserved only through 24 wavebands, and the method has the advantages of being high in robustness, high in accuracy and high in accuracy. According to the provided swarm intelligence-driven heterogeneous model collaborative optimization architecture, prediction errors are compressed to an agronomy applicable standard, and a technical case giving consideration to mechanism interpretability and landing feasibility is established for hyperspectral agricultural condition monitoring.

The invention discloses a working condition self-adaptive excavator pressure sensor soft measurement method, which comprises the following steps of: firstly, constructing a working condition self-adaptive variable screening framework, and integrating causal analysis, working condition sensing variable selection and a soft attention gate mechanism to realize accurate screening and weighting of key characteristic variables; further, designing an adaptive Kalman filtering algorithm, and dynamically adjusting parameters according to working conditions to suppress noise interference; on the basis, a CGM hybrid network fusing a convolutional neural network, a gating circulation unit and a multi-head attention mechanism is built, so that the time sequence feature extraction capability is enhanced; and furthermore, through an improved IPOA algorithm, CGM network hyper-parameters are optimized, and an IPOA-CGM high-precision pressure prediction model is constructed to predict a pressure value. The method can effectively meet the control requirements under various working conditions, provides reliable soft measurement signals when the sensor fails, improves the robustness and safety of the system, and has good engineering application value.

The invention provides an adult postoperative shivering prediction system and method based on machine learning, and the system comprises a data processing module which is used for obtaining a multi-source data variable of an adult surgical patient, carrying out the feature type division and missing value processing of the multi-source data variable, and obtaining a processed multi-source data variable; the variable screening module is used for screening the processed multi-source data variables by using LASSO regression to obtain key prediction variables; and the shiver prediction module is used for performing adult postoperative shiver prediction by using the key prediction variable and a shiver prediction model based on limit gradient lifting to obtain a prediction result. According to the technical scheme, scientific reference is provided for clinically selecting the optimal prediction model, and the application range of machine learning in postoperative complication prediction in the medical field is expanded.

The invention provides a plateau traumatic acute kidney injury risk prediction method and device, and the method comprises the steps: carrying out the interpolation and standardization post-processing of missing values in a training set after the elimination of abnormal values and a verification set after the elimination of abnormal values through employing a preset multiple interpolation method; performing preliminary variable screening on the processed training set by adopting a standard LASSO regression method to obtain variables after preliminary screening; carrying out secondary variable screening on the preliminarily screened variables by adopting a self-adaptive LASSO regression method to obtain secondarily screened variables; correspondingly constructing a plurality of initial acute kidney injury risk prediction models according to the secondarily screened variables and a plurality of preset machine learning models; comparing the prediction accuracy of the plurality of initial acute kidney injury risk prediction models through the processed verification set, and screening out a target acute kidney injury risk prediction model; and performing risk prediction through the target acute kidney injury risk prediction model to obtain an acute kidney injury risk prediction result.

The invention discloses a necrotizing enterocolitis prediction method based on machine learning. The method comprises the following steps: collecting multi-source clinical data of neonatal cases; carrying out systematic data preprocessing on all the collected data; carrying out feature engineering and variable screening by adopting a plurality of feature selection algorithms, and randomly dividing finally obtained feature data into a training set and a test set according to a proportion; and after model training is completed, performance verification is carried out on each candidate model by adopting an independent test set, and comprehensive evaluation is carried out in combination with multiple indexes derived from the classification confusion matrix. According to the method, the problem of limited feature dimension of a traditional model can be broken through, and the information utilization rate is improved through multi-modal data fusion; the method can solve the problem that a traditional linear model is difficult to capture a nonlinear relation between variables, achieves the interaction recognition between complex features, provides an intelligent auxiliary decision-making tool for intensive care of newborns, reduces the disease progress risk, and reduces the case fatality rate and the complication occurrence rate.

The invention discloses a transformer substation waterlogging situation water level prediction method and device, electronic equipment, a medium and a product. The method comprises the following steps: inputting multi-source monitoring variable time series data into a preset dynamic variable screening model to obtain monitoring variable fusion features; the preset dynamic variable screening model is constructed based on a variable selection network; inputting the monitoring variable fusion features into a preset time sequence modeling model to obtain multi-scale fusion features; the preset time sequence modeling model is constructed based on a bidirectional gating circulation unit; inputting the multi-scale fusion features into a preset adaptive prediction model to obtain a water level prediction result; the preset adaptive prediction model is constructed based on a feedforward neural network. According to the scheme, the adaptability to a dynamic scene and the prediction robustness are enhanced by dynamically selecting key variables which have obvious influence on waterlogging condition and water level prediction; the characterization capability of the complex sequential relation is improved through multi-scale modeling, so that the precision and timeliness of water level prediction are improved.

The invention provides a desulfurization and denitrification system dynamic collaborative optimization method based on a physical constraint heterogeneous agent, which comprises the following steps: carrying out preprocessing and variable screening on original operation data to obtain a core operation variable set; performing optimization constraint and normalization operation on the original operation data to obtain a normalized data set so as to construct a desulfurization and denitrification collaborative prediction agent model; embedding a running sample into a regeneration kernel Hilbert space, carrying out dynamic environment detection, and then carrying out nucleation selection and genetic matching to obtain a diversity enhanced population; and performing dynamic collaborative multi-objective optimization by using the desulfurization and denitrification collaborative prediction agent model and the environmental change judgment result, continuously updating the Pareto solution set, and outputting an optimal operation strategy. According to the method, on the premise that emission constraint and equipment safety requirements are met, physically consistent high-precision agent modeling and perceptible and traceable dynamic working conditions can be carried out, and efficient, stable and economical operation of a desulfurization and denitrification system in a wide load interval is achieved.

The invention relates to the technical field of application of artificial intelligence in assisted reproduction technology, and discloses an assisted reproduction multi-node clinical decision-making method based on machine learning. The method comprises the following steps: collecting sample data, obtaining sample features, and carrying out conversion and interpolation on the sample features; performing predictive variable screening on the sample features by using a Spearman correlation coefficient; dividing the data of the complete sample into a training set, a test set and a verification set according to a sample proportion of 8: 1: 1, taking a predicted variable obtained by screening as an independent variable in the training set, and taking accumulated live birth within 2 years after single egg taking as a dependent variable; respectively constructing prediction models in different decisions of four stages of a controlled ovarian stimulation scheme, a gonadotropin initiation amount and the like, and correcting to obtain prediction values; and carrying out hyper-parameter adjustment by using the test set, and comparing live birth outcomes of the crowds which accord with and do not accord with the recommendation by using the verification set. According to the method, corresponding cumulative live yield prediction can be provided for various feasible schemes, and a single optimal path is not recommended.

The present application relates to the field of short-term wind power generation prediction, and discloses a few-sample wind power prediction method fusing mechanism migration modeling, step 1) wind power generation data preprocessing and dataset division; step 2) MIC characteristic variable screening; step 3) defining characteristic variables and labels on the source domain and the target domain; step 4) establishing a KAN mechanism fusion model on the source domain; step 5) source domain pre-training and physical guided migration to the target domain; step 6) model training, prediction and model performance evaluation on the target domain; the present application introduces a migration strategy of freezing the physical layer, migrates the trained physical network layer in the source domain to the target domain, and only fine-tunes the prediction module, so as to realize wind power modeling under the condition of low samples; the strategy fully retains the physical feature expression ability in the source domain, significantly improves the prediction accuracy and stability under the condition of small sample learning of the target domain, and shows good migration generalization ability.

The invention relates to the crossing field of public health and geographic information technology, and discloses a plague risk assessment method and system combined with a MaxEnt model, and the method comprises the steps: obtaining a multi-source environment variable, and carrying out the standardization; reducing a colinear variable through variance threshold preliminary screening, a Pearson's correlation coefficient and plague point significance test; performing variable importance sorting by using recursive feature elimination and a support vector machine, and dynamically determining an optimal variable subset; and inputting a MaxEnt model to train ecological niche probability distribution, and dividing risk levels based on historical occurrence point probability quantiles. The system comprises corresponding function modules. According to the system, through a three-stage automatic variable screening mechanism, the generalization ability, interpretability and prediction precision of the model are remarkably improved, and reliable support is provided for accurate prevention and control of plague.

The application belongs to the technical field of wind turbines, and relates to a residual life probability prediction method for key components of large wind turbines. After obtaining wind turbine SCADA data through a data acquisition and monitoring control system, variable screening and preprocessing are first performed, then different periodic changes of the wind turbine SCADA data are preliminarily found through Fourier transform, then multi-dimensional time series data are intercepted according to different periods, then parallel processing is performed through an LSTM neural network, time characteristics are deeply extracted, then output vectors of multiple LSTM neural networks are spliced together and residual life prediction results are obtained through a linear output layer, finally, an uncertainty quantification method based on linear regression and kernel density estimation is used to quantize the uncertainty of the original residual life prediction results, thereby providing a more robust reference for wind turbine maintenance decisions. Compared with other methods, the method can effectively improve the prediction accuracy and reliability of the residual life of wind turbine components.

The invention discloses a method and a system for constructing a marginal region lymphoma early progress prediction model based on Lasso-Cox regression, and relates to the field of biological medicines. The method solves the problems that an existing tool cannot meet cross-subtype and cross-site stable generalization and accurate recognition clinical requirements of early-stage high-risk groups, and the like, and comprises the steps that 1, historical data is obtained to serve as a training set and a verification set, and the historical data is marginal region lymphoma early-stage sample data; step 2, carrying out multiple interpolation on missing sample data by calling a rice packet; and step 3, using Lasso-Cox regression to construct a marginal region lymphoma progress prediction model, and the construction method is as follows: based on data after multiple interpolation, using a Lasso regression model to carry out variable screening, introducing an L1 regularization term into the regression model of a target variable, realizing coefficient sparsification by minimizing a loss function with a penalty term, and obtaining a prediction result of the marginal region lymphoma progress prediction model. And an optimal regularization parameter = 0.102 is adopted as a working point for screening, and a more accurate risk layering tool is provided for prognosis evaluation.

The invention discloses a time-resolved metal material multi-mechanical parameter synchronous detection device and method, and belongs to the technical field of metal material mechanical property detection. The method comprises the following steps: 1, configuring sample spectrum acquisition by using a time-resolved metal material multi-mechanical parameter synchronous detection device; 2, respectively constructing a spectrum matrix Mn under different delay time; 3, sequentially carrying out saturation removal, abnormity removal, background removal and normalization on the spectrum matrix to form a sample initial feature vector; 4, respectively acquiring training set feature matrixes under different delay times, and forming training set feature input of the prediction model by adopting a variable screening and recombination mode; 5, performing multi-fold cross validation by using the randomly selected part of samples to adjust hyper-parameters so as to obtain a prediction model; the trained prediction model is used for metal mechanical parameter detection; compared with the prior art, the technical problem of low accuracy in mechanical parameter detection of the metal material is solved by adopting the time-resolved spectrum.

The invention discloses a light symptom SFTS risk prediction method and system based on machine learning, and the method comprises the steps: obtaining light symptom SFTS clinical data, carrying out the preprocessing, carrying out the variable screening of the preprocessed data, determining an optimal regularization parameter, obtaining a key prediction factor, building a Cox proportional risk scoring model of two time points based on the key prediction factor, and carrying out the calculation of the Cox proportional risk scoring model. Calculating individual risk scores and constructing a column graph, determining risk levels based on a risk layering threshold value, realizing risk prediction, finally generating a visual heat map according to the combination of the key prediction factors, marking critical disease risk probabilities under different key prediction factor combinations, and completing risk prediction of different key prediction factor combinations. The objective of the method for constructing the risk prediction model of the mild SFTS based on machine learning is to realize accurate risk prediction of early-stage SFTS critical disease progress through double-time-point modeling, dynamic risk layering and visualization of a heat map.

The application discloses a variable screening method and device, a nonvolatile storage medium and a processor. The method comprises the following steps: obtaining independent variables for evaluating a target variable, wherein the target variable corresponds to one or more independent variables; using a preset Pearson correlation coefficient model to evaluate the linear relationship between the independent variables and the target variable, and determining the linear correlation degree; in the case that the target variable corresponds to multiple independent variables with a linear correlation degree higher than a preset correlation degree threshold, selecting the independent variable with the highest linear correlation degree as a sample variable of the target variable, wherein the sample variable and the target variable are used as training data for training a target prediction model, and the target prediction model is used for analyzing the independent variables to determine a prediction variable. The application solves the technical problem of low efficiency in determining a consumer portrait due to the inability to screen the independent variables for determining the consumer portrait.

The invention discloses a variable screening method and device, a storage medium and electronic equipment, and the method comprises the steps: determining a variable contribution value corresponding to a plurality of variable combinations related to risk prediction in a business data set, the variable contribution value being a contribution value of a first variable to a risk prediction result corresponding to a second variable; extracting a plurality of first variables in the plurality of variable combinations, and dividing the plurality of first variables into a plurality of variable intervals according to a preset business range; counting a first proportion corresponding to a positive value variable contribution value and a second proportion corresponding to a negative value variable contribution value in the plurality of variable intervals, and determining a first variable set based on a size relationship between the first proportion and the second proportion; calculating the interaction degree between different variables in the first variable set and the integrating degree between the different variables and preset business logic, and screening the first variable set based on the interaction degree and the integrating degree to obtain a second variable set.

The application discloses a method and equipment for predicting individual phenotypes based on human whole genome genotypes, and the method comprises the following steps: obtaining haplotypes in all regions of the whole genome of each individual, converting the haplotypes into functional genome parameters, and selecting at most one representative parameter; quantitatively analyzing the correlation between each region and individual phenotypes by using a regression model; obtaining a phenotype prediction value of each gene by combining a variable screening model and a first gradient ascent network; and inputting the phenotype prediction value of each gene into a second gradient ascent network for integration to generate a final prediction result. The application can integrate information from various functional genome changes at a single gene level, avoids the limitation of linear correlation of isolated analysis of each gene site, and finally integrates the prediction values of all genes by using a gradient ascent network to fully reflect the nonlinear relationship between genes, thereby improving the prediction accuracy and having important significance for the prediction of various individual phenotypes.