9 results about "Regression result" patented technology

The invention discloses a thunderstorm gale regression analysis method and system based on multi-stream parameters, and relates to the technical field of data processing. The method comprises the following steps: collecting multi-stream parameters in real time, and constructing a real-time multi-stream parameter sequence; mapping the real-time multi-flow-parameter sequence to a pre-constructed multi-flow-parameter coupling sensitivity field model, and calculating to generate a real-time convection sensitivity field; inputting the real-time convection sensitivity field into the physical consistency constraint model to generate real-time structured convection characteristics; main regression analysis under multi-flow parameter regression is executed, and a main regression result is established; and after the real-time local disturbance index is calculated, disturbance residual compensation of the main regression result is triggered, dynamic weight adjustment is carried out on the main regression result after residual compensation by utilizing seasonal phase characteristics, and a thunderstorm gale regression analysis result is output. The technical problem that thunderstorm gale prediction accuracy is insufficient in the prior art is solved, and the technical effect of improving thunderstorm gale regression analysis result accuracy is achieved.

The invention discloses a semi-supervised regression method and device based on pseudo-label heterovariance hypothesis, and relates to the technical field of machine learning. Iteratively executing the model training step until a stop condition is met; constructing a double-layer optimization framework to perform joint optimization on parameters of the regression model and parameters of the pseudo-label uncertainty estimation network; predicting a pseudo label of the sample by using the regression model, performing heterovariance modeling on the pseudo label of the unmarked sample to obtain lower layer loss, and updating parameters of the regression model based on the lower layer loss; obtaining upper layer loss based on the updated regression model, and updating parameters of the pseudo label uncertainty estimation network based on the upper layer loss; inputting a to-be-processed regression task into the trained regression model, and outputting a corresponding regression result; and selecting different evaluation indexes according to different sub-data sets to evaluate the accuracy of a regression result. The problem of how to avoid how to accurately measure the uncertainty of the false label by purely depending on consistency regularization is solved.

This application provides a method, apparatus, and electronic device for training an object detection model. The method includes: acquiring sampled positive samples, each positive sample containing a bounding box corresponding to a sample image; classifying the positive samples using a classification branch of a pre-built object detection model to obtain a class probability for each positive sample; regressing the positive samples using a regression branch to obtain a regression result for each positive sample; calculating the intersection-over-union ratio (IoU) for each positive sample based on the regression result and the bounding box; determining a weight for each positive sample based on the class probability and / or the IoU; adjusting the smoothed regression loss in the object detection model using the weights to obtain an adjusted regression loss; and training the object detection model using the adjusted regression loss. This application improves the training effect of the object detection model, resulting in higher accuracy.

The invention relates to the technical field of text processing, in particular to a text processing method for financial cases. The financial case text processing method comprises the following steps: acquiring case description of a predetermined financial case, and extracting case features through a language model; obtaining a macroscopic digital index of the financial case; determining alternative case features and alternative digital indexes based on the macroscopic digital indexes and the correlation between the case features and the processing targets; and performing regression processing based on the alternative case features and the alternative digital indexes to obtain a regression result. Therefore, by fusing the case features and the macroscopic index features of the financial cases and utilizing the feature engineering method and the predetermined regression model to carry out regression processing, prediction and analysis of different dimensions of the financial cases can be realized.

The invention relates to the technical field of vehicle motor system performance development, in particular to a parameter identification method and device for a motor stator mode, equipment and a medium. According to the method, N to-be-identified parameters are determined according to anisotropic material parameters of a motor stator modal, so that the N to-be-identified parameters are used for calculating the frequency of the corresponding modal, M groups of experimental data are determined according to a preset experimental design method, and finite element modal simulation is carried out on the motor stator modal by using each group of experimental data, so that the modal of the motor stator is simulated. Each group of experimental data is used for fitting the agent model, so that the situation that a regression result is insensitive or too sensitive to some coefficients due to the fact that the difference between the upper limit span and the lower limit span of each parameter is too large can be avoided, and the accuracy of coefficient items in the target agent model is improved; therefore, the optimization function is constructed according to the target agent model, and the accuracy of the optimized parameter value is improved when the parameter value of the to-be-identified parameter is optimized and adjusted.

This invention discloses a method and system for thunderstorm gale regression analysis based on multiple convection parameters, belonging to the field of data processing technology. The method includes: real-time acquisition of multiple convection parameters to construct a real-time multiple convection parameter sequence; mapping the real-time multiple convection parameter sequence to a pre-constructed multiple convection parameter coupled sensitivity field model to calculate and generate a real-time convection sensitivity field; inputting the real-time convection sensitivity field into a physical consistency constraint model to generate real-time structured convection characteristics; performing a master regression analysis under multiple convection parameter regression to establish the master regression result; after calculating the real-time local disturbance index, triggering disturbance residual compensation of the master regression result, and dynamically adjusting the weights of the residual-compensated master regression result using seasonal phase characteristics, outputting the thunderstorm gale regression analysis result. This solves the technical problem of insufficient accuracy in thunderstorm gale prediction in existing technologies, achieving the technical effect of improving the accuracy of thunderstorm gale regression analysis results.

The adaptive outlier identification method for dam deformation monitoring data disclosed in this invention mainly involves introducing the Baseline Management System (BMS) method to optimize the key influencing factors in the dam deformation monitoring model. Then, a robust regression model for dam deformation is established using the optimized set of key influencing factors. Combined with the Longest Trigger Score (LTS) estimation method, robust regression analysis is performed on the dam deformation monitoring data. This allows for simultaneous regression modeling, outlier identification, and structural deformation prediction without the need for data preprocessing. This adaptive outlier identification method can adaptively overcome the misleading effect of outliers on regression, enhancing the significance of the regression results and effectively improving the accuracy of data prediction.

The application relates to a method for analyzing track angle monitoring service parameter errors of non-direct flight tracks, wherein track angle data of specific tracks with extremely small data amounts are subjected to regression, distribution fitting and other analysis and calculation, in the regression stage, the obtained prediction value is taken as the actual track angle value of the track, the original data is taken as the observation value, and the difference between the actual value and the observation value is calculated; in the distribution fitting stage, the regression result is subjected to track angle error distribution fitting analysis, the mean value of the track angle error is taken as the track angle error of the track section, and thus the track angle monitoring service parameter error is obtained. In the case that the track angle data amount is extremely small, the accuracy of the calculation result is improved.

The present invention discloses a drilling cycle prediction method based on a random forest regression model, which relates to the field of drilling data processing technology and includes the following steps: S1, collecting and screening drilling data to obtain standard drilling data; S2, classifying and processing the standard drilling data to obtain classification results; S3, obtaining a random forest-based regression result based on the classification results, and using the regression result to predict the drilling cycle. The present invention can meet pre-drilling prediction and real-time parameter adjustment requirements as needed, continuously supplementing training samples to optimize the model, and can improve the model's adaptability to the operational characteristics of the study area. The random forest regression prediction model is of great significance for subsequent drilling plan planning and cost assessment, and has good promotion value.