55 results about "Statistical significance" patented technology

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 relates to a training method of a respiratory tract infection disease progress and prognosis prediction model. The training method comprises the following steps: extracting mRNA from peripheral blood of a target patient, and carrying out transcriptome sequencing to obtain a sequencing result; based on the ferroptosis related gene set, comparing ferroptosis score differences of two groups of patients with community-acquired pneumonia and sepsis, and screening corresponding ferroptosis related genes with statistical significance from a sequencing result; screening out genes meeting preset conditions from the ferroptosis related genes based on LASSO regression; and establishing an RTI clinical outcome prediction model through logistic regression by taking whether the patient is sepsis or not as an outcome dichotomy variable and taking the screened gene expression quantity as a prediction variable. According to the invention, after the prediction model is subjected to machine learning screening such as LASSO and the like, the core feature with the highest prediction value is reserved, so that the risk of over-fitting of the model on training data is reduced.

The invention discloses a single cell clustering statistics significance test method and system based on manifold learning, and the method comprises the following steps: obtaining original single cell data, carrying out the standardization processing of the original single cell data, and removing the deviation and noise in the data; performing dimension reduction on the standardized data by using a manifold learning method; dividing the dimension reduction data based on a pre-calculated clustering label, and performing hierarchical clustering on the divided data to obtain a hierarchical tree; hypothesis testing is carried out step by step from the bottom to the top of the hierarchical tree by adopting a Monte Carlo simulation method, and whether the clusters are combined or not is determined according to a hypothesis testing result; and finally, taking the combined hierarchical tree as a single cell clustering correction result. According to the method, the reliability and the accuracy of the clustering statistical test can be improved, and unnecessary group splitting and misjudgment can be avoided.

The invention provides a closed watershed lake water level evolution prediction method based on multi-source data, and relates to the technical field of hydrological resources. The method comprises the following steps: firstly, collecting natural environment and social economic multi-source data of a closed drainage basin, and constructing a normalized multi-source feature input matrix; carrying out total element analysis on the features and extracting a nonlinear response threshold value; and meanwhile, Mantel statistical test is executed on the feature matrix to obtain a statistical correlation coefficient. And judging the characteristics which simultaneously meet the statistical significance and mechanism threshold conditions as real causal driven factors, and eliminating false related factors which are statistical significance but have no physical threshold. And finally, reconstructing a prediction model based on the screened causal-driven feature set, and outputting a water level evolution prediction value. Through'statistics-mechanism 'dual verification, interference of false correlation is effectively eliminated, the problem that a traditional black box model cannot distinguish causal and coincident trends is solved, and physical interpretability and robustness of water level prediction are remarkably improved.

The invention discloses an ecological system service dominant factor identification method. The method comprises the following steps: 1) establishing a basic ecological system service influence factor database; 2) selecting key evaluation indexes for evaluating the target area and quantifying the key evaluation indexes; 3) constructing a multiple linear regression model, and introducing a nonlinear term correction model; 4) performing primary screening on influence factors in the correction model based on a variance expansion factor VIF; 5) further screening the influence factors by using optimal subset feature screening, and establishing a corresponding influence factor index system; and 6) carrying out local regression analysis by adopting an improved multi-scale geographically weighted regression MGWR model, and identifying the ecosystem service dominant factor according to an obtained regression coefficient value. The method disclosed by the invention not only can reveal the dominant factors of ecosystem services in different areas and on different scales, but also ensures that the identified dominant factors have statistical significance through significance test on the regression coefficient, and enhances the credibility and practicability of an analysis result.

The invention discloses a multi-technology fusion prediction method based on hypergraph modeling, and relates to the technical field of computers. According to the method, hyperedge weights are corrected by introducing a probability-based zero model so as to eliminate deviations caused by patent number differences among different technical fields, and then a multi-technology fusion mode with statistical significance is identified. Further converting a multi-technology fusion prediction problem into a hyperedge prediction problem in a hypergraph, calculating an average similarity among technologies in a technology combination by adopting a random walk algorithm with restart, and establishing a mapping relation between a feature and a hyperedge formation probability through a logistic regression model by taking the average similarity as the feature; therefore, prediction of the technology fusion probability is realized. According to the method, the fusion trend of three or more technical fields can be effectively identified and predicted, the method is more suitable for the analysis requirement of current multi-technology cross-field cross fusion, and data support and decision basis are provided for technical innovation layout and industrial policy planning.

The embodiment of the invention discloses an ecological system simplification processing method and device based on breakpoint regression, and the method comprises the steps: setting a function component simplification sequence for a target ecological system; according to the function component simplification sequence, gradually removing corresponding function components from the target ecological system, and monitoring and collecting system performance indexes of the target ecological system; using a breakpoint regression method to identify a critical simplified point where the system performance index is suddenly changed; and according to the statistical significance index, the effect quantity and the system structure difference of the target ecological system before and after the critical simplification point, checking whether the target ecological system reaches an unsimplification degree or not. According to the method, the breakpoint regression method is introduced into the non-simplification research of the ecological system, the non-simplification critical state of the ecological system can be objectively and accurately recognized, and a scientific decision basis is provided for the fields of ecological system protection, ecological system design optimization and the like.

The invention discloses an adaptive enhanced dynamic graph contrast learning multivariate time sequence classification method, which is oriented to a multivariate time sequence classification task under the condition of few labels, and comprises the following steps: 1) an adaptive enhanced strategy; 2) dynamic graph comparison learning; and 3) carrying out two-stage combined training. According to the method, in a preprocessing stage, trend-periodic decomposition and statistical significance test based on random permutation are combined to judge the strength of time sequence dependence, and an enhanced recommendation strategy matched with data characteristics is generated; and dynamic graph comparison characterization learning and a two-stage training mechanism are introduced, so that the classification performance and the model stability are improved while unlabeled data are fully utilized.

The application discloses a kind of high-rise building group health monitoring method and system based on timing InSAR, belong to building health monitoring technical field.The method includes: obtaining multi-temporal SAR image, extracting scattering point deformation time series;Identify high-rise building group and group according to spatial position;To single building vertically divided height layer;Calculate the average deformation sequence of each height layer, the seasonal deformation parameters of the same height layer of the same building are counted, and a group deformation reference model is constructed;Abnormal height layer is identified based on dynamic threshold and statistical significance test;Map the anomaly to the corresponding building, and if there are multiple abnormal height layers in the same building, the structure abnormal risk is determined.The method does not require external temperature data, can effectively distinguish seasonal deformation and structural anomaly, and improve monitoring accuracy and automation level.

To provide an estimation and control method for frequency errors of extreme events.SOLUTION: The present invention proposes a frequency shift functional and a frequency shift functional estimation formula based on probability estimation, and constructs a method for reversely solving a confidence frequency shift value epsilon according to the frequency shift functional to improve a frequency distribution function. The present invention is suitable for estimating and controlling errors of occurrence frequency prediction of extreme events in various frequency statistical models by using hydrological observation data under probability statistical significance, so that the prediction accuracy and reliability of hydrological extreme events such as storm surge, rainstorm flood, tsunami, and the like are improved, risks caused by extreme natural disasters are reduced, and loss of life and property that may be caused is reduced.SELECTED DRAWING: Figure 1

The invention relates to the technical field of data processing, and discloses a report generation method based on statistical significance weighting and causal constraint, which comprises the following steps: receiving structured data; performing cleaning, normalization, feature engineering and basic statistical analysis on the data to generate an insight data set; executing a statistical significance test, and performing preliminary screening on the insight data according to a preset significance threshold value; based on the screened insight data; performing dynamic weighting and priority ranking on all contents to be reported to generate a report core abstract; and taking the generated report core abstract and the causal constraint mark as input, driving a natural language generation model, and finally outputting a report. According to the method, the risk of excessively interpreting random fluctuation is eliminated from the source by screening insight data; through the causal relationship graph, the accuracy and business guidance of the graph can be improved; the causal constraint marks are used for declaring to ensure that all conclusions are verified by causal reasoning, and the logic accuracy of the report is guaranteed.

The application relates to the technical field of data fusion, in particular to a multi-dimensional data feature fusion method for enterprises combined with AI technology. The method takes the attribute data of any dimension as the attribute to be analyzed for the enterprises in the same industry, determines the statistical significance of the attribute to be analyzed and the risk value through a K-S test method, compares the attribute to be analyzed and the risk value to determine the risk deviation consistency of the attribute to be analyzed, combines the statistical significance and the risk deviation consistency to determine the importance of risk assessment, adds the attribute data to a candidate set according to the size order of the importance, determines the recognition degree of the attribute data added to the candidate set, selects the candidate set corresponding to the maximum recognition degree as the final candidate set, and takes the attribute data in the final candidate set as the feasibility attribute data to perform weighted fusion on the feasibility attribute data according to the recognition degree. The application only performs feature fusion on the selected attribute, avoids data redundancy, and improves the data fusion effect.

The invention discloses a medical term analysis method based on a knowledge graph technology, and the method comprises the following steps: obtaining medical standard literatures of different time versions, and constructing a medical term knowledge graph sequence containing time dimension information; based on the medical term knowledge graph sequence, calculating semantic representation changes of the target medical term at different time points to obtain a semantic evolution path of the target medical term; acquiring clinical disease course data related to the target medical term; based on a statistical significance test method, calculating the support degree of the clinical disease course data to emerging semantics in the semantic evolution path, and generating a quantitative verification result; and dynamically updating and optimizing the medical term knowledge graph sequence according to the quantitative verification result. According to the method, prospective active perception and accurate verification of emerging semantics are realized, and the maintenance mode of the knowledge graph can be converted into active optimization from post passive updating, so that potential medical risks caused by indefinite term evolution are effectively reduced.

The invention discloses a method for predicting heart failure based on left ventricular ejection fraction, which develops and trains a primitive model to automatically provide help and evaluate the level of clinical decision based on LVEF heart failure. The training model is prevented from using any or biased clinical variables, and the following two steps are ensured: investigating the statistical significance of each variable to distinguish the three categories; secondly, a novel dimension reduction technology is adopted to visually observe the characterization of the optimal variable in the radial direction, and each LVEF-based HF category is separated; and on the basis, the performance of the developed model is trained, the importance of the most important clinical variables is discussed, and the significance of the clinical variables is explained in detail based on the application of LVEF-based deep learning in HF analysis, so that patient data of three heart failure types are distinguished.

The present invention provides a method for evaluating earthquake precursor prediction effectiveness and optimizing prediction parameters. The method preprocesses earthquake precursor signals and earthquake catalog data, performs a grid search on parameter combinations to be optimized to generate multiple different parameter combinations, uses a preset prediction strategy to predict the preprocessed data based on the multiple different parameter combinations, generates a Molchan plot corresponding to each parameter combination based on the prediction results corresponding to each parameter combination, the number of earthquake catalog samples, and a preset probability distribution, and determines the optimal parameter combination from the multiple different parameter combinations based on the first surface fraction value between the actual prediction curve and the confidence interval in each Molchan plot. The method can eliminate the impact of differences in the number of earthquake catalog samples on earthquake precursor prediction effectiveness evaluation, improve the optimization efficiency of multiple parameters, and enhance the statistical significance of earthquake precursor prediction effectiveness evaluation results.

This invention discloses a method for determining the root density threshold under optimal erosion reduction effect, relating to the fields of soil and water conservation and vegetation engineering technology. The method includes: obtaining sediment yield and root density sequences; constructing a fitting model of sediment yield variation with root density using a nonlinear fitting algorithm; detecting abrupt change points using various hydrological abrupt change detection algorithms; constructing an evaluation model including consistency, statistical significance, effect size, and repeatability indicators; calculating a comprehensive score for each candidate abrupt change point; selecting the candidate abrupt change point with the highest comprehensive score as the optimal root density threshold; and verifying the optimal root density threshold using a multiple verification algorithm, outputting the threshold result that passes verification. This method overcomes the shortcomings of relying on experience and lacking unified standards in determining the optimal erosion reduction root density threshold through algorithmic processes and systematic evaluation and verification, achieving automated and objective determination of the root density threshold.

The invention discloses a system and a method for determining early cancer risk influence factors. The method comprises the following steps: taking blood routine, blood biochemical and tumor marker data of a cancer group; determining a peripheral blood-derived inflammatory marker for the group of cancers; comparing the tumor marker data and the peripheral blood-derived inflammatory marker with corresponding data of a corresponding normal group, and determining a key abnormal factor with statistical significance; constructing a Lasso regression model based on all the key abnormal factors, and screening out a first-level influence factor from all the key abnormal factors according to the selected target parameters; a single-factor logistic regression analysis and parameter estimation method is utilized to screen out second-level influence factors from all the first-level influence factors; and a multi-factor logistic regression analysis and parameter estimation method is utilized to determine a significant impact factor from all the secondary impact factors. According to the method, the obvious influence factors on the early-stage cancer can be more effectively determined, and the accuracy and the reliability of cancer risk early warning are improved.

The invention discloses a method for predicting the material demand of a deep learning model based on swarm intelligence, and aims to solve the problems that a conventional hyper-parameter optimization method is low in efficiency and is liable to fall into local optimum. The method comprises the following steps: firstly, constructing a deep learning model containing the number of layers, the number of hidden nodes, a learning rate and regularization parameters, and defining an optimization interval of each hyper-parameter; secondly, designing an improved sparrow search algorithm (ISSA), enhancing global search capability through Cauchy variation disturbance, balancing exploration and development of a refraction reverse learning strategy, and adjusting a self-adaptive early warning value to improve convergence stability so as to optimize a hyper-parameter combination; further taking a verification set RMSE as a fitness target, combining with parallelization calculation to accelerate model performance evaluation, and screening optimal hyper-parameter configuration; and finally, predicting material demand data by using the optimized model, performing transverse comparison with LSTM, ARIMA, GRU, GCN and other models, and verifying the effectiveness of the model through multiple indexes such as RMSE, MAPE and R. Experiments show that the ISSA optimization method provided by the invention significantly reduces prediction errors, the optimized model has higher accuracy and robustness in a complex material demand scene, and meanwhile, the reliability of performance improvement is confirmed through statistical significance test.

Methods are disclosed for determining a block error rate (BLER) and determining if the BLER is significantly above, below or on BLER target. The method includes determining a pair of thresholds including an upper threshold and a lower threshold for a predetermined confidence level and determining a threshold in total number of blocks received. The method further includes determining a Required Compensation BLER from a number of bad blocks, a total number of blocks received and a maximum total number of blocks to receive for a purpose of statistical significance; comparing the Required Compensation BLER against the upper and lower threshold to determine if the BLER is significantly above the BLER target, below the BLER target or inconclusive; and, if the threshold in total number of blocks received is reached, then the inconclusive result is reinterpreted as the BLER being on target.

The invention relates to the technical field of data fusion, in particular to an enterprise multi-dimensional data feature fusion method combined with an AI technology. According to the method, for enterprises in the same industry, attribute data of any dimension serves as attributes to be analyzed, and statistical significance of the attributes to be analyzed and risk values is determined through a K-S test method; comparing the to-be-analyzed attribute with the risk value, and determining the risk deviation consistency of the to-be-analyzed attribute; the importance of risk assessment is determined in combination with statistical significance and risk deviation consistency; according to the importance sequence, adding the attribute data into the alternative set, and determining the identification degree of adding the attribute data into the alternative set; selecting the alternative set corresponding to the maximum identification degree as a final alternative set; and taking the attribute data in the final alternative set as feasibility attribute data, and performing weighted fusion on the feasibility attribute data according to the identification degree. According to the method, only the selected attributes are subjected to feature fusion, data redundancy is avoided, and the data fusion effect is improved.

The invention discloses a bridge influence line identification method and a damage verification method based on a physical information neural network, and relates to the field of bridge health monitoring, and the method comprises the steps: 1, extracting an initial influence line through employing bridge physical parameters and actual measurement dynamic response, and constructing a multi-dimensional feature project; 2, constructing a PINN model based on intelligent boundary learning, and adjusting natural learning bridge endpoint physical constraints through dynamic weights to realize high-precision prediction of influence lines; and 3, establishing a self-adaptive threshold value based on health sample statistical characteristics, and realizing intelligent judgment of bridge damage in combination with multiple verification logics (simple threshold value, statistical significance and regional continuity). According to the method, the noise immunity and the boundary accuracy of influence line identification are effectively improved, and the bridge damage can be quickly and automatically positioned and the severity can be evaluated.

The invention provides a cardiac postoperative infection risk prediction method and system, and the method comprises the following steps: S1, data collection: collecting clinical data and image data of a patient, the image data comprising spleen artery hemodynamic parameters obtained through stomach and abdomen ultrasound; s2, feature preliminary screening: performing Logistic single-factor regression analysis on the clinical data and the image data, and screening out candidate features according to statistical significance; and S3, analysis and calculation: carrying out Lasso regression analysis on the candidate features, determining key features and weight coefficients thereof, and calculating to obtain an infection risk probability value after the cardiac surgery. The infection risk prediction method and prediction system provided by the invention can specifically predict and identify infected high-risk patients so as to guide clinical intervention measures to be taken as soon as possible.

The present invention discloses a satellite data full-space assimilation method based on weather scale and weather background dependence, comprising the following steps: (1) reading a background error covariance matrix containing information of N sample members, calculating and defining a scale classification operator; (2) reading the background error covariance matrix containing information of N sample members again, and constructing an adaptive partitioned background error covariance of variable scale dependence; (3) calculating a background error covariance based on large-scale variables; (4) calculating a background error covariance based on small-scale variables; (5) reading information containing background error covariance and its diagnostic quantity, satellite observation information, satellite observation error information and model background field, and inputting them into a variational assimilation system; (6) obtaining a model initial field with the best statistical significance; the present invention improves the accuracy of subsequent numerical forecasts.

The invention discloses a method for establishing an AB experiment evaluation system, and the method is characterized in that the method comprises the steps: building characteristics based on a target scene, building a characteristic layer, and distributing the characteristics to the characteristic layer; the characteristic is an event which is supposed to be related to a user click behavior; determining related experiment events of the target scene, and distributing the related experiment events to different experiment layers; the related experiment event is an event related to a user click behavior; dividing the user data into orthogonal shunts according to the experimental layer and then entering the characteristic layer, determining a control group and an experimental group of an AB experiment according to characteristics, and recording results of the AB experiment; and checking the significance of the AB experiment according to the result of the AB experiment. According to the method for establishing the AB experiment evaluation system, a flow-limiting access flow distribution mode is adopted, and different flow distribution factors are selected according to different experiments, namely user data are divided into a characteristic layer and multiple experiment layers. The method comprises the following steps: independently and randomly sampling from different experiment layers, mixing with a characteristic layer, randomly sampling again, establishing an indifferent experiment group and a control group as the basis of AB experiment shunting configuration, and when the result of the AB experiment has no statistical significance, considering that the divided experiment group and the control group have no difference.

The invention discloses a unimolecule conductivity measurement data consistency evaluation method, electronic equipment and a storage medium. The method comprises the following steps: firstly, carrying out histogram statistics on two groups of single-molecule conductance data to extract a peak position characteristic sample; then executing Welch's t test and double-single-side test (TOST), and simultaneously judging the statistical significance and the physical equivalence of the data under a preset physical equivalence limit; for the non-normal distribution sample, further carrying out robust judgment by adopting Mann-Whitney U test and Hodges-Lehmann estimation, and carrying out robust judgment on the sample according to the robust judgment; and finally, constructing a Bayesian probability model based on Student-t distribution, and calculating a posterior probability of a difference parameter falling into an actual equivalent interval (ROPE) through MCMC sampling. According to the method, the frequency acknowledgement and the Bayesian inference are fused, the problem that the physical consistency cannot be confirmed only by depending on value test in the prior art is solved, and a quantitative repeatability evaluation standard is provided for single molecule sensing and junction cracking experiments.

This invention belongs to the field of runoff prediction technology, and specifically discloses a runoff prediction method and system based on statistical significance constraints. The method includes: extracting features from hydrological and meteorological time-series data of the target prediction section and at least one upstream hydrological station to obtain candidate features; performing statistical significance testing and multiple hypothesis testing on the candidate features, combined with information theory correlation screening and feature redundancy screening, to obtain a feature subset; constructing a runoff prediction model including a variable selection subnetwork, a time lag gating subnetwork, and a main prediction network; training the runoff prediction model using the feature subset and historical runoff increments of the target prediction section; and realizing runoff prediction based on the trained runoff prediction model. This invention can improve the reliability of feature inputs and the stability of runoff prediction output.

The invention relates to supervised machine learning of a method for carrying out a technical process (TPR), in which a data pool (DP) containing data records (x;y) with input data and output data describing a correct process result is created, the method is trained in a training phase (TP), wherein process parameters of the method are varied during repeated process executions, in a validation phase (VP) the trained method is checked by comparing the output data calculated with the trained method using the input data with the output data describing the correct process result and as a result of the comparison an actual failure probability (Ptrue) for the method is calculated.It is proposed that all data sets (x; y) in the data pool (DP) are made available both as training data sets (TD) and as validation data sets (VD). For the training phase (TP), an empirical failure probability (Pemp) is specified that is at most as large as a predetermined target failure probability (PFD). The validation phase (VP) is initiated after it has been determined in the training phase (TP) that the empirical failure probability (Pemp) is not exceeded. Furthermore, it is proposed that a statistical significance (α) of the result of the validation phase (VP) is taken into account by specifying a corrected failure probability (Pval) in the validation phase (VP) for comparison with the actual failure probability (Ptrue) that is lower than the predetermined target failure probability (PFD).

The invention discloses a spatial aggregation detection method for large-scale disease data, and the method comprises the following specific steps: A, dividing a to-be-detected region into a plurality of geographic units, carrying out the modeling of the to-be-detected region into an undirected graph structure based on the geographic units, and creating an adjacent matrix; b, constructing a candidate scanning window set of each geographic unit based on an adjacent constraint control mechanism and a low-risk region elimination mechanism; c, searching a maximum likelihood cluster in the scanning window set based on a spatial search mechanism of the scanning window and taking the maximum likelihood cluster as a candidate result; and D, carrying out Monte Carlo simulation test and statistical significance analysis on the maximum likelihood cluster in the candidate results obtained in the step C, and selecting the aggregation region conforming to a set significance threshold as the aggregation region. According to the method, the accuracy and the detection efficiency of disease space aggregation detection under large-scale geographic data and fine scale can be effectively improved.

The invention relates to the technical field of bed separation grouting, in particular to an evaluation method for determining a bed separation grouting range based on intelligent drilling parameter inversion lithologic change. The core of the method is to introduce a hidden structure dip angle parameter, establish an information magnitude correlation model of a strong transformation ratio coefficient, a deformation ratio coefficient and a stratum structure, distinguish lithologic anomaly caused by the structure and characteristics of a rock stratum, and solve the problem of inversion deviation of a traditional method. Dynamic data comparison is formed through two times of drilling and calculation before and after grouting, the statistical significance of interval estimation quantification difference is combined, evaluation is converted into a quantifiable confidence level index, and the limitation of experience judgment is broken through. And directly outputting a filling effect conclusion and an adjustment suggestion based on the comparison of the confidence level and the discrimination threshold, optimizing the decision-making efficiency, and reducing the cost and the construction period. And meanwhile, drilling multi-dimensional parameter values are deeply excavated, and upgrading of data from collection and storage to decision support is achieved.

The present application discloses a method, device and storage medium for long-term monitoring of epileptic seizures, and selects a single EEG channel and four EEG features. The present application filters and segmentally pre-processes the EEG signals of all channels pre-measured by the patient, uses the power spectral density function (PSD) parameterization technology for feature extraction, characterizes the non-periodic component with two parameters, the offset and the exponent, selects the two largest powers and their corresponding center frequencies in the periodic component, and the original PSD corresponding to the center frequencies of the two periodic components is recorded as the total feature power; selects the four key features most relevant to the seizure event; selects the most representative channel according to the statistical significance P value of the four key parameterized features during the seizure period and the inter-seizure period, and performs data enhancement on the EEG data of the channel; finally, performs automatic detection of epileptic seizures based on the support vector machine classifier. The present application does not require pre-specified frequency sub-bands, and can adaptively select representative channels for specific patients. It has good tolerance for individual differences, can achieve consistent performance for each patient, and has strong robustness.