12 results about "Selection bias" patented technology

The invention discloses an electric power line fault early warning method based on big data analysis, and relates to the technical field of electric power fault early warning, and the method comprises the steps: collecting electric power line operation monitoring data, generating a working condition feature vector and an operation and maintenance strategy feature vector, constructing operation and maintenance big data, and dividing the operation and maintenance big data into region data sets according to region numbers; training a physical mechanism characterization function, constructing an early warning basic model, generating an original early warning probability, performing federal parameter aggregation, and obtaining a global physical mechanism characterization function and a global early warning basic model; generating a region selection bias correction factor, correcting an original early warning probability, and generating an updated global physical mechanism characterization function and an updated global early warning basic model; and obtaining a real-time working condition feature vector and a real-time operation and maintenance strategy feature vector, calculating a real-time original early warning probability, obtaining a real-time correction early warning probability, and generating a fault early warning level. Self-adaptive compensation of regional sample distribution differences is realized, and the fault early warning reliability is improved.

A system and method for improving predictive modeling of patient outcomes by transforming ctDNA-linked genomic data into engineered pathway and network features and training machine-learning models that explicitly incorporate ctDNA status and amount while controlling selection bias. By broadening the cohort to include both ctDNA-positive and ctDNA-negative patients, harmonizing ctDNA across assays, and iteratively reducing features using Shapley-based importance, the approach improves predictive accuracy, calibration, and clinical relevance.

The application relates to the technical field of electric digital data processing, in particular to a data storage method, a system and a data center platform. The method comprises the following steps: determining a target selection bias of a target user for parking at a target parking space; the selection bias represents the possibility of the user selecting any parking space for parking; grouping the parking spaces based on parking time points, determining the time point consistency degree and the parking space distance between the parking spaces in a target group; determining a target parking range where the target parking space is located according to the target selection bias, the time point consistency degree and the parking space distance; determining a storage priority coefficient by using the target selection bias of the target parking space and the average selection bias of the target parking range, and determining a storage priority according to the storage priority coefficient. Through the data storage method, the adaptive ability of the user parking data storage is effectively improved, and the storage effect is enhanced.

The invention discloses a large model analysis system and method for endometrial cancer data management, and relates to the technical field of data large model analysis, and the method comprises the steps: determining the total amount of sample data; each group of sample data comprises characteristic data information of the patient and a judgment result of molecular typing data determined by adopting an immunohistochemical method; the features are divided based on the feature data information of the patient, the coincidence rate under each feature division is calculated, and a plurality of coincidence subintervals are formed based on the coincidence rates; and constructing a training model, forming the probability that the coincidence rate is in different coincidence subintervals based on different feature data information, forming a recommendation decision, and feeding back the recommendation decision to an administrator port. The method can solve the problem of significant subjectivity existing in a traditional mode of selecting a detection method only depending on the personal experience of doctors, establishes an objective decision standard through data driving, and reduces the selection deviation caused by the experience difference of the doctors.

The invention belongs to the technical field of vascular access catheterization evaluation, and discloses an intelligent evaluation method and device before vascular access catheterization and a storage medium, and the method comprises the steps: firstly obtaining medical data of three features of a patient's treatment scheme, a vascular condition and a host risk, and constructing a comprehensive feature vector through numeralization and standardization processing; inputting the adaptive probability vector into a pre-trained multi-classification gradient lifting decision tree model to obtain an initial adaptive probability vector and performing probability calibration; correcting the calibrated probability based on a clinical rule set containing a hard constraint rule and a soft penalty rule, and generating a final recommendation probability vector; and finally, outputting a recommendation list sorted according to the probability and key feature contribution information. Quantitative and accurate evaluation of the vascular access is achieved, selection deviation caused by experience dependence is avoided, complication risks are reduced, clinical decision-making efficiency and safety are improved, the decision-making process is transparent and explainable, and clinical practical application requirements are met.

The invention discloses an advertisement feedback optimization method and device based on machine learning, and relates to the technical field of digital advertisements. The method comprises the following steps of: firstly, constructing a causal graph containing user characteristics, advertisement exposure, click and conversion, eliminating selection deviation caused by confusion variables by utilizing a dual machine learning model, and accurately estimating a condition average processing effect of advertisement putting; secondly, performing multi-contact attribution in combination with a conditional average processing effect and a Shapley value to generate an initial putting strategy, and dynamically updating the strategy by using real-time data through an incremental causal forest algorithm; and finally, screening high conditional average processing effect material features, generating a new idea by using a generative adversarial network, and feeding back the new idea to a delivery engine. According to the method, the problems of correlation deviation, inaccurate attribution and delayed creative optimization in traditional advertisement putting are solved, and the advertisement putting accuracy, the real-time response capability and the return on investment are remarkably improved.

The invention relates to the technical field of wisdom education, and discloses a computerized adaptive test depolarization method based on selective mixing, which comprises the following steps: constructing an interaction data set based on obtained historical answer records of examinees, and encoding to obtain answer state vectors; obtaining an ability vector representing the cognitive state of the examinee according to the answer state vector; defining a deviation alignment sample, an unbiased sample and a deviation conflict sample; selective mixing is carried out, and diversified problem characterization is generated by carrying out linear interpolation on problem parameters, so that deviation conflict samples are expanded, and a synthetic sample set is obtained; constructing final training loss including empirical loss and synthetic sample loss, and optimizing the selection network to relieve selection deviation; through cross-attribute retrieval and selective mixing, a small amount of deviation conflict samples can be fully utilized and expanded to cope with cognitive diagnosis challenges caused by unbalanced data distribution, so that more accurate cognitive diagnosis is realized, and more excellent test experience is brought to examinees.

This invention discloses a method and apparatus for predicting the clinical intervention effect based on a diffusion-cross counterfactual regression network, comprising: a patient feature dataset construction module, a clinical intervention effect prediction model construction module, a model optimization and training module, and a clinical intervention effect prediction model application module. By constructing a patient feature dataset and building and training a drug clinical efficacy prediction model based on a diffusion-cross counterfactual regression network, this invention effectively solves the problems of potential confounding factors, selection bias, and sample imbalance in clinical observation data. This approach significantly improves the accuracy and reliability of predicting individual patient intervention effects and can accurately distinguish between patient subgroups with and without effective interventions, thereby providing decision support for personalized medicine.

The invention relates to the technical field of artificial intelligence, and discloses a self-adaptive weight language model adjustment method based on environment feedback, which comprises the following steps: firstly, establishing topological mapping of expert nodes and communication channel groups; secondly, channel-level aggregation and capacity normalization are carried out on the collected flow, and a dominant congestion distribution vector representing a network bottleneck structure is extracted by using an iterative algorithm; then, calculating the projection proportion of the current load on the dominant congestion distribution vector to determine the locking strength; and finally, generating a suppression coefficient and a smoothing parameter based on the locking strength, which are respectively used for adjusting expert selection bias and output scoring distribution of the router. Through cross-layer feedback and two-dimensional nonlinear control, intrinsic mode locking of network congestion is effectively broken, and communication delay of distributed reasoning is reduced.

Embodiments of the invention provide a data processing model training method and apparatus, and a data processing method and apparatus. The data processing model training method comprises the steps of determining a first conditional probability corresponding to first experimental data and a second conditional probability corresponding to second experimental data; training according to a target calibration weight calculated according to the first conditional probability and the second conditional probability to obtain a calibration estimator; inputting the target confusion data and the corresponding observation result into a calibration estimator to obtain a first calibration weight, and inputting the target confusion data and the estimation result into the calibration estimator to obtain a second calibration weight; processing the estimation result according to the two calibration weights to obtain a calibration result, inputting the target confusion data and the target intervention data into an initial data processing model to obtain a prediction result, and training according to the prediction result and the calibration result to obtain a target data processing model; the problem of selection deviation in survival analysis is solved by calibrating the estimator and adjusting the weight.

The invention relates to the technical field of tunnel micro-seismic signal identification and data feature optimization, and discloses a tunnel micro-seismic signal classification model construction method based on feature contribution degree screening, and the construction method comprises the steps: obtaining an original elastic wave signal, extracting multi-domain features, and constructing an initial full-feature data set; training an XGBoost base classifier, a Light GBM base classifier and a CatBoost base classifier on the basis of the data set; calculating the contribution degree of each feature, and obtaining a comprehensive value through a normalization and weighted fusion mechanism; setting an adaptive threshold to screen key features, and constructing an optimized key feature set; and obtaining a key feature data set based on the set reconstruction data, and finally training a final classification model. According to the method, by fusing multi-model feature contribution degree results, the defect of single model feature selection deviation is overcome, and effective compression and optimization of a high-dimensional feature space are realized. The constructed classification model is simple in structure and low in calculation cost, high recognition precision is kept, and the generalization ability and the real-time application performance of the model are remarkably improved.

The invention discloses an active domain adaptation method based on large model uncertainty, and belongs to the technical field of deep learning. According to the method, screening and labeling of target domain samples are assisted through a large language model, and progressive adaptation under the cross-domain condition is achieved. Firstly, in order to reduce sample selection deviation and reduce large model calling overhead, a distance weighting strategy of an energy function is introduced on the basis of a nearest centroid classifier, and a target domain sample with potential labeling value is preliminarily screened out; next, aiming at a special cue word in the data set style design field, calling a large model in combination with a meta cue word to carry out multi-round labeling on a screening sample, generating a plurality of candidate results in each round, and then carrying out uncertainty modeling on the candidate results through Bayesian aggregation to determine a final unique labeling result; and finally, training a classifier by mixing the labeled target domain sample and the source domain sample to obtain a model with good performance on the target domain. According to the method, dependence on artificial experts in a traditional active domain adaptation task is reduced, active domain adaptation is achieved by combining energy distance screening and uncertainty modeling, and an automatic and efficient solution is provided for active domain adaptation related tasks.