33results about How to "Quantitative uncertainty" patented technology

The invention relates to the field of the mode identification, and discloses a method for carrying out sentiment classification on a Chinese comment text on the basis of the ensemble learning and a system on the basis of the method. The method comprises the following steps of: a, acquiring the Chinese comment text from a webpage and carrying out preprocessing on the Chinese comment text; b, sequentially training a multi-classifier system in parallel; c, classifying the comment text to be classified by a base classifier and converting a classification output into an intuitionistic fuzzy number; and d, combining a weight and a guide variable of the base classifier, fusing the sentiment tendency of the comment text to be classified and making a classification decision. The method and the system have the following advantages that the training and classifying speed is ultrahigh; a sequence learning strategy is adopted, so that newly developed vocabularies can be found conveniently and the requirement on a corpus can be lowered; and the classification accuracy is improved by the ensemble learning, and thus, the system on the basis of the method can support the management or purchase decision more satisfactorily.

A hard disk life prediction method based on backpropagation Bayesian deep learning, which optimizes the SMART parameters of the hard disk, screens out SMART parameters with predictive performance, and then performs standardized processing, and then divides the standardized data into sequence samples ; Extract the historical time series information and future time series features of the hard disk; input the historical time series information and future time series features into the linear fully connected layer, and adjust the activation function and learning rate to complete the construction of the remaining life prediction network; generate the remaining life according to the distribution parameters after training The weight parameters and bias parameters of the life prediction network are used to predict the remaining life; the invention integrates the multi-dimensional features of the hard disk, makes full use of the long-term dependence of the hard disk data, extracts the timing information in the hard disk data, and uses the deep learning framework. While achieving high-precision remaining life prediction, quantify the uncertainty of the prediction, and output the prediction confidence with higher guiding significance for hard disk maintenance decisions.

The invention relates to the field of pattern recognition, and discloses an unbalanced data classification method based on intuitive fuzzy integration and a system based on the method. The method comprises the following steps of: a) cleaning original data, and classifying original point-of-sale (POS) class samples according to intra-class positions to generate POS class artificial samples; b) training a base classifier by using different sample sets of inter-class approximate balance; c) converting the classification output equal utility of the base classifier into an intuitive fuzzy matrix; and d) integrating samples to be classified into the membership and the non-membership of the POS class and the negative (NEG) class by combining the weight of the base classifier, and making a classification decision. The invention has the advantages that: over learning is avoided by integrating over sampling and under sampling; the training samples of the base classifier are different, so that the difference of the base classifier is ensured; the base classifier is not specifically limited, so the method has good expandability; the intuitive fuzzy reasoning method quantitatively describes the uncertainty in classification so as to improve the performance of integrated learning; therefore, the system based on the method can better support the medical diagnosis decision and the like.

The invention discloses a photoelectric probability density prediction method based on B-spline quantile regression, and the method comprises the steps: 1, collecting photoelectric data, carrying outthe normalization of the photoelectric data, and dividing historical photoelectric data into a training set and a test set; 2, constructing a B-spline quantile model, and calculating parameters of theB-spline quantile regression model by utilizing the training set data; and 3, substituting the test set data into the B spline quantile model to obtain predicted values under different quantiles, andrealizing photoelectric probability density prediction by applying kernel density estimation. The prediction precision of photovoltaic power generation can be improved, and the uncertainty of a prediction result is comprehensively measured, so that a reliable basis is provided for safely and stably connecting the photovoltaic power generation into a power grid.

The invention discloses a TVF-EMD-MCQRNN load probability prediction method based on fuzzy C-means clustering, which includes: 1. grouping according to a set time interval after preprocessing the power load and its influencing factors; 2. grouping The data set is divided into training set and test set, and the fuzzy C-means clustering method is used to cluster each group of training set and test set; Test to obtain a series of conditional quantiles of various types of sample data at different quantile points; 4 sum the conditional quantiles of each type of sample data at the same quantile point to obtain various types of prediction results, thereby realizing Probability Density Prediction. The invention can improve the accuracy of load prediction, thereby providing more comprehensive and effective load information for the operation planning of the power system.

The present invention proposes a high-speed rail track response prediction method based on sparse Bayesian width learning. The method includes linear and nonlinear feature extraction for input temperature field variables, and maximum output layer weights for hidden layer neuron nodes. Posterior estimation, prediction of structural response output, and preliminary assessment of track structure state, etc. The present invention adopts the sparse Bayesian breadth learning method to mine the correlation relationship of the data of the high-speed rail track monitoring system, and can effectively avoid the over-fitting problem of the regression prediction by sparsely solving the weight value w reflecting the relationship between the data variables, and has relatively good performance. High prediction accuracy, efficient calculation speed and loose equipment hardware requirements, so as to realize the mining of the relationship between temperature load and structural strain hidden in a large amount of monitoring data, and timely discover the evolution of monitoring data models as a way to judge the service life of track structures The basis for the abnormal status.

The invention relates to a method for revising structural model parameters based on a frequency response function, comprising the following steps: collecting time history data and time history response data, introducing the multivariate circular symmetric proportional distribution theorem to derive the probability density function and covariance of the measured frequency response function Matrix; introduce the prediction error and the parameters to be corrected to obtain the covariance matrix containing the parameters to be corrected; according to the determinant of the matrix and the inversion theorem, the probability density function of the frequency response function under single-point excitation is obtained; according to the maximum likelihood principle, The maximum likelihood function expressed in the form of the maximum likelihood function and the logarithmic maximum likelihood function is obtained; according to Bayes' theorem, the posterior probability density function of the random variable is obtained; the posterior probability density function is then expressed as a logarithm The form of the likelihood function is to obtain the objective function. The invention quantifies the uncertainty of the correction parameters, improves the calculation accuracy of the correction parameters, and realizes the correction of the structure finite element model.

The invention discloses a comprehensive decision-making method for a power system based on double-layer fuzzy optimization, which belongs to the field of power system management. The invention proposes that in power system planning, power production and environmental management are layered for decision-making, and the fuzzy uncertainty of economic parameters in the system is considered, and a double-layer fuzzy programming method is used to solve the problem. In the process of solving, decision makers at different levels control different decision variables and optimize their respective goals; the decision makers at the upper level make decisions first, and the lower levels react according to the decision results of the upper level. Therefore, the optimization method disclosed in the present invention obtains the optimal power system planning scheme by balancing the interests of the two layers, and the optimization results disclosed in the present invention can provide technical support for power system planning, comprehensive utilization of resources and environmental protection in the research area.

The present invention relates to a method for predicting the static parameter distribution of reservoir geological modeling based on the nearest neighbor neural network. Coordinates and corresponding static parameter values; S200 uses the nearest neighbor algorithm to find the adjacent wells of each well; S300 establishes and trains the neural network model; S400 uses the optimal neural network model obtained from training to predict Static parametric distribution of points in unknown space. This method makes full use of the excellent ability of the neural network to approximate complex nonlinear functions, and can dig deep into the nonlinear distribution relationship of static parameters in space, which conforms to the complex characteristics of reservoir geology, can improve the accuracy of spatial interpolation, and can also be realized through multiple stochastic , to quantify the uncertainty of spatial interpolation and improve the accuracy of static parameter distribution predictions.