442 results about "Independent predictor" patented technology

The systems and methods described herein utilize a probabilistic modeling framework for reverse engineering an ensemble of causal models, from data and then forward simulating the ensemble of models to analyze and predict the behavior of the network. In certain embodiments, the systems and methods described herein include data-driven techniques for developing causal models for biological networks. Causal network models include computational representations of the causal relationships between independent variables such as a compound of interest and dependent variables such as measured DNA alterations, changes in mRNA, protein, and metabolites to phenotypic readouts of efficacy and toxicity.

Methods and systems for designing an experiment using a computer to determine whether the experiment is a true experiment are described. These approaches allow a user who is unsophisticated in the complexities of true experimental design to design and deploy an experiment that produces substantially confound-free results and can be used to determine and quantify any causal relationship between independent and dependent variables. The computer may select one or more independent and/or dependent variables of the experiment or may assist the user in selection of independent and/or dependent variables. Formation of control and treatment groups, randomization and/or blocking to reduce the effects of confounding variables may be performed by the computer with or without input from the user.

An evidence based cost modeling and predictive analysis system, and an incentives based plan to reduce healthcare costs are disclosed. An analytics system may generate incremental expenditures among overweight and obese individuals, predictive forecasts of future medical costs, and predictive forecast of cost reduction based on financial incentives to recipients. The forecasts may include statistical trends, prevalence of diseases based on body mass index, and medical evidence associated with specific illnesses. A computer based program may process and analyze dependent and independent variables in electronically stored information (for example insurance, health and medical records). A health insurance provider may provide an annual rebate on paid premiums to recipients based on a qualifying annual BMI as an incentive. The recipients may receive the rebates in a qualified health reimbursement account (HRA) managed by the recipients towards future healthcare related expenditures.

The invention provides a classifier construction method, and a method of prediction classification. The classifier construction method includes the steps: acquiring a training data set of a pluralityof training samples, wherein the training data set includes attribute information and classification information; extracting the attribute characteristic from the attribute information; taking the attribute characteristic as an initial independent variable, and taking the corresponding classification information as an initial dependent variable to perform at least one round of model training, wherein at least candidate models take part in each round of model training; and taking the combination with the minimum error rate in the each round of model training as a classifier which has completedtraining. The classifier construction method, and the method of prediction classification perform at least one round of model training on the basis of at least two candidate models to obtain a classifier and can predict classification of target samples according to the trained classifier, thus being able to avoid the problem that a single classification method is low in prediction precision and prediction accuracy, and being both higher in precision and accuracy of prediction.

The present invention relates to a steady state optimization method incorporating dynamic variance correction for dynamic variations of both independent variables and dependent variables of a dynamic system. The dynamic variance correction is based on measured variance of the variables and a weighing factor for each of the variables. The dynamic variance correction offers an effective method of dynamic violations avoidance of controlled variables for a model predictive controller without having to constantly adjust the tuning weights in response to changing dynamical conditions.

The invention discloses an intelligent early warning method for dam safety monitoring data. The method comprises the steps of early warning model establishment, threshold value setting and mutual feedback type early warning. Gross error identification and gross error processing are carried out, model sample data quality is improved, according to the monitoring items, independent variable relevance, historical monitoring data quantity and historical monitoring data distribution, different early warning models and indexes are established, including a stepwise regression model, a correlation vector machine model and a gray system model; the established models can reflect the relationship between the independent variable and the dependent variable more truly and are wide in application range,according to a measuring instrument, measuring point attributes, a threshold value, an early warning model and indexes, real-time early warning is carried out on monitoring data, monitoring instrumentabnormity early warning is sent to monitoring personnel, or dam safety early warning is sent to dam safety management personnel, experts with professional knowledge and rich experience are not needed, the workload is small, the early warning speed is high, and the early warning result is more accurate and reliable.

The invention discloses a social network link prediction method adopting knowledge graph embedding and a time convolution network. The method comprises the following steps: S1, processing original social sample data, extracting phrases and tuples of independent variables related to the phrases, constructing structured event tuples, linking the structured event tuples to a knowledge graph, constructing sub-graphs from the knowledge graph, and extracting event embedding vectors; S2, expressing the network structure of the social network by using an adjacency matrix, and fusing an event embeddingvector and a network adjacency matrix in a vector form; and S3, establishing a link prediction model based on the improved time convolution network, taking a fusion vector of the event embedding vector and the network adjacency matrix as input of the prediction model, and obtaining an optimal model through iterative training so as to predict a social network link. According to the invention, theprediction precision of the social network link can be improved.

The invention relates to a non-invasive serological scoring model for hepatic fibrosis and a design method thereof. The model includes an input module, a processing module and an output module. The input module receives three independent predictors including blood platelet count, laminin and procollagen type 3. The processing module is connected with the input module and evaluates the situation ofhepatic fibrosis based on the independent predictors. The output module is connected with the processing module and outputs the score finally obtained by the module. The model can more accurately evaluate the situation of severe hepatic fibrosis for patients with chronic hepatitis. Compared with other models, the model only involves the three independent predictors including PLT, LN and PIIINP.The contained serological indicators are fewer, the accuracy and the practicability are higher, the model is more concise to use, popularization in clinical treatment and primary hospitals is facilitated, the probability of invasive examination is thus reduced, and the risk and burden of the patients are reduced.

The present disclosure provides for analyzing expenditures by category. A file including expenditure data relating to spent monies is identified. The expenditure data from the file is extracted. An independent and a dependent variable relating to the expenditure data are determined. The expenditure data is graphed based on the determined variables. Conclusions are displayed based on the expenditure data.

A method of data management for optimizing the patient outcome from the provision of external counterpulsation (ECP) therapy is described. This method describes a process by which sets of dynamic cardiopulmonary dependent variables are measured during steady-state conditions, displayed, and translated into quantitative and qualitative measurements while the independent variables of ECP, cuff inflation duration and cuff inflation pressure settings of ECP systems, are altered by a physician. In combination with visual observation and computer-assisted ranking of the dependent variables, a physician can utilize the resulting information to render decisions on the optimal choice of the independent variables. The method will enable physicians to collect, view, track and manage complicated data using well-understood visualization techniques to better understand the consequences, acutely and chronically, of their therapeutic actions in general, and of their provision of ECP therapy in particular.

The invention belongs to the technical field of underground tunnel engineering construction and provides a method for quantitatively predicting the cutter abrasion quality of the heading machine cutter during composite stratum heading. On the basis of tool mass measurement and corresponding tool and tunnel face contact point trajectory calculation, mass line loss indexes MLI of tools at differentpositions are subjected to classification statistics according to strata. After the MLI is normalized into equivalent mass line loss indexes EMLI, a universal prediction model for the equivalent massline loss indexes of homogeneous strata is obtained. The MLI and the corresponding EMLI are clear in physical significance. The equivalent mass line loss index universal prediction model considers both the shareability of independent variables and the prediction precision. The cutter wear quality prediction method based on the equivalent mass line loss index general prediction model is provided. The calculation process is simple and clear. The prediction precision is high. Based on engineering investigation data and construction data, the method is reasonable, high in practicability and beneficial to quantitatively predicting the tool abrasion state, scientifically and reasonably arranging bin opening and tool changing and improving the tunneling efficiency.

The invention discloses a machine tool thermal error modeling method based on deep learning. The method comprises the following steps of obtaining temperature data of a machine tool main shaft and carrying out normalization processing; calculating correlation between each measuring point and thermal errors of a main axis at three directions through a partial correlation coefficient method, and selecting m measuring points with high correlation as temperature critical measuring points; establishing a SAE network and initializing a network parameter, and taking temperature data of the temperature critical measuring points as independent variables and inputting into the SAE network to extract a temperature characteristic of the temperature data; and taking the temperature characteristic as the independent variable, taking corresponding thermal error data as a dependent variable and inputting into a GA-BP neural network to carry out training, and carrying out thermal error prediction. Themethod has advantages that prediction precision is high, robustness is good, a thermal error change trend of the machine tool can be effectively estimated and so on.

An estimator for producing values for dependent variables of streaming objects based on values of independent variables of the objects, as well for continuously tuning the estimator based on any objects received with pre-populated values for the dependent variables.

The invention discloses an intelligent wastewater monitoring method based on complex network multivariate online regression, which comprises the following steps: collecting historical data including independent variables and dependent variables; performing normalization processing on the collected historical data to obtain a normalization model, and training the normalization model to obtain a trained normalization model; taking the independent variable as the input of the normalized model after training, carrying out the online learning of the normalized model after training, and updating thestate of the model in real time; performing reverse normalization processing on the output dependent variable to obtain a predicted dependent variable, and further regulating and controlling the wastewater treatment system. The complex network multivariate online regression method constructed by the invention solves the problem of poor generalization performance of deep learning on long and shortsequence regression, can be used for water quality parameter prediction, realizes intelligent water quality monitoring of a wastewater treatment system, and promotes efficient and stable operation ofthe wastewater treatment system.

The invention discloses a high-speed pantograph structure parameter optimization method which comprises the following steps of 1, determining the pantograph working performance influence factors, and respectively constructing a kinematics optimization model, a statics optimization model, a dynamics optimization model and a control science optimization model of a pantograph; 2, establishing an integrated optimization model according to the model obtained in the step 1; 3, solving the integrated optimization model obtained in the step 2 to obtain an optimization result of an independent variable; and 4, establishing a three-dimensional model of the pantograph according to the optimization value of the variable, and carrying out finite element analysis to complete the optimization of the pantograph. According to the present invention, on the basis of a complete theoretical model, the interaction of various factors is fully considered, and compared with the serial design, the design result better meets the requirements of the practical engineering.

The invention provides a fault influence factor quantitative analysis method based on a high voltage switch and can acquire accurate relationship between fault influence factors and corresponding fault types. The method comprises steps (1) initial data is acquired according to operation parameter data and fault type data after data classification; (2), standard processing on the initial data is carried out; (3), the operation parameter data taken as an independent variable and the fault type data taken as a dependent variable are introduced to regression analysis to acquire an optimal high voltage switch fault influence factor quantitative analysis model; (4), regression diagnosis of a Logistic regression equation of the model is carried out, if the model is qualified, the progress turns to a step (5); if not qualified, a secondary optimal high voltage switch fault influence factor quantitative analysis model of the step (3) is utilized to carry out regression diagnosis till the modelis qualified, the progress turns to the step (5); (5), according to the qualified optimal high voltage switch fault influence factor quantitative analysis model of the step (4), quantitative analysison the Logistic regression equation is carried out to acquire high voltage switch fault influence factors.

This disclosure describes systems and methods for using an estimator to produce values for dependent variables of streaming objects based on values of independent variables of the objects. The systems and methods may include continuously tuning the estimator based on any objects received with pre-populated values for the dependent variables.

The invention provides a fault diagnosis method and device for a wind generating set. The fault diagnosis method includes the following steps: acquiring actual operation data of the wind generating set at each sampling time point in the current time period, wherein the actual operation data comprises a first group of actual operation data related to a first group of operation parameters and a second group of actual operation data related to a second group of operation parameters, the first group of operation parameters are independent variables related to the fault, and the second group of operation parameters are dependent variables of the first group of operation parameters; dividing the current time period into a plurality of sub-time periods based on the obtained first group of actualoperation data in the current time period, the difference between the first group of actual operation data in two sub-time periods which are continuous in time satisfying a first preset condition; anddetermining whether the wind generating set has a fault or not based on the second group of actual operation data in the plurality of sub-time periods.

The invention provides a power transformation equipment fault rate prediction method and system, equipment and a readable storage medium. The power transformation equipment fault rate prediction method comprises the steps: collecting the historical fault rate statistical data of power transformation equipment, and building a Weibull distribution fault rate function with the fault rate as a dependent variable and the time as an independent variable; collecting substation equipment state evaluation data, and establishing an equipment health index-based fault rate model by taking the fault rate as a dependent variable and the substation equipment health index as an independent variable; and configuring a function relationship between the equipment health index and the average operation time,performing simulation training on the comprehensive prediction model by taking historical statistical data of the power transformation equipment as a sample, solving to-be-estimated parameters, and finally obtaining the comprehensive prediction model. According to the power transformation equipment fault rate prediction method, data information such as online monitoring data, daily maintenance data and routine test data of the power transformation equipment is combined, and the equipment state condition is comprehensively reflected, and the development requirement of state maintenance of a current power system is met, and the fault rate of the power transformation equipment can be accurately predicted.

A method is provided that includes generating a visual environment for interactive development of a machine learning (ML) model. The method includes accessing observations of data each of which includes values of independent variables and a dependent variable, and performing an interactive exploratory data analysis (EDA) of the values of a set of the independent variables. The method includes performing an interactive feature construction and selection based on the interactive EDA, and in which select independent variables are selected as or transformed into a set of features for use in building a ML model to predict the dependent variable. The method includes building the ML model using a ML algorithm, the set of features, and a training set produced from the set of features and observations of the data. And the method includes outputting the ML model for deployment to predict the dependent variable for additional observations of the data.

Embodiments of the invention disclose a sales volume data prediction method and device, and related equipment. The method comprises the steps of obtaining sample data by electronic equipment; whereinthe sample data comprises store sales volume data and feature data corresponding to the store; performing feature extraction on the feature data and the store sales volume data to obtain a feature sample; taking the store sales volume data as a dependent variable of the first training sample, and taking the feature sample as an independent variable of the first training sample to form a first training sample; training the clustering model based on the first training sample to obtain N clustering cluster samples; training the regression model based on the N clustering cluster samples to obtainN regression models; wherein the trained clustering model and regression model are used for carrying out sales data prediction on the to-be-predicted store. By adopting the embodiment of the invention, the clustering category of the predicted store can be determined, and the regression model corresponding to the clustering category is selected for prediction, so that the problem of inaccurate prediction caused by insufficient model training is solved, and the accuracy of business prediction is improved.

The invention discloses a satellite precipitation space downscaling optimal regression window screening method and device, and the method comprises the steps: obtaining digital elevation data and normalized vegetation index data of a first spatial resolution as independent variables, carrying out the resampling of a second spatial resolution, and obtaining satellite precipitation data of the second spatial resolution as dependent variables; constructing three regression models of a global window, a local window and a pixel-by-pixel change local window according to the independent variable andthe dependent variable of the second spatial resolution, and screening an optimal local window and an optimal pixel-by-pixel change local window; and according to the regression model and the independent variable data of the first spatial resolution, obtaining rainfall prediction data of the first spatial resolution under the three typical regression windows, and determining an optimal regressionwindow through verification of actually measured rainfall station rainfall data. The method can effectively screen the optimal regression window of satellite rainfall downscaling, and is simple and easy to operate and high in prediction precision.

The invention provides a method for forecasting forest pest occurrence degree which uses occurrence index number to measure occurrence of forest pest, combines the occurrence degree with an occurrence area, and can accurately reflect occurrence conditions of the forest pest. Occurrence degree classification is conducted according to the occurrence index number and based on data within a large region; meteorological data is not directly used in forecasting, and change of meteorological factors is used as independent variable. Therefore, occurrence of the forest pest in different geographical scales is unified, and a forecasting model can be used for forecasting not only in small regions (city or county) but also in large regions (province or whole nation).

The invention relates to a target device identification method, an electronic device and a medium, and the method comprises the steps: S1, obtaining the information of a plurality of sample devices, constructing a training set, wherein the sample devices comprise a target device and a non-target device; S2, preprocessing the sample data of the training set based on a preset first algorithm framework; S3, acquiring hyper-parameters of a preset model based on a preset second algorithm framework, wherein the preset model comprises a plurality of sub-models; S4, performing model training based ona preset first algorithm framework, the preprocessed sample data of the training set and hyper-parameters of a preset model to obtain a device classification model; and S5, obtaining an independent variable feature vector corresponding to a to-be-tested device, inputting the independent variable feature vector into the device classification model to obtain a classification prediction value, judging whether the classification prediction value is greater than a preset classification threshold or not, and if so, determining the to-be-tested device as a target device. According to the invention, the target device identification efficiency of the scene with device information changing in real time is improved.