22 results about "Uncertainty estimation" patented technology

A computer-implemented method utilizing a continuous discrete recurrent Kalman network, wherein the method includes receiving, at an encoder, an input from one or more sensors, wherein the input includes one or more time series data associating data at one or more points in time; outputting, to a Kalman filter, a latent observation and uncertainty estimate in response to the input at the encoder; determining a latent state prior and latent state posterior utilizing the Kalman filter; and outputting, via a decoder, a filtered observation utilizing at least the latent state posterior.

One embodiment of the present invention provides a marker pose uncertainty estimation method. The method comprises the steps of: receiving a single image that includes a reference marker, and deriving a corner position for the reference marker on the basis of a pixel intensity variation of the reference marker; deriving corner position uncertainty for the corner position on the basis of the pixel intensity variation, and deriving the position and the pose of the reference marker on the basis of the corner position; and deriving the position and the pose uncertainty of the reference marker on the basis of the corner position and the corner position uncertainty.

The present application relates to a multi-modal data inconsistency detection and cleaning method and system based on adaptive energy guidance, belonging to the field of artificial intelligence and data mining technology. The present application proposes a multi-modal data inconsistency detection method based on adaptive energy guidance. First, a multi-expert uncertainty estimation framework is constructed to effectively mine multi-dimensional modal semantic correlation information such as prediction uncertainty, cross-modal consistency and teacher-student difference. Second, according to the multi-expert evaluation signal, an adaptive energy scoring mechanism is designed to effectively aggregate and quantify the sample modal consistency by constructing a comprehensive energy function and a weighted function based on Boltzmann distribution. Finally, a dynamic data cleaning and screening strategy based on energy score is constructed to improve the quality of the data set and the robustness of the downstream model, and better achieve the purpose of data service.

This invention relates to the field of computer vision and discloses a facial expression recognition method based on enhanced action units-guided causal inference. The method includes extracting high-dimensional semantic feature maps from the original image; performing reparameterized sampling to output a sequence of local visual feature blocks; calculating the geometric relationship between the feature sequence and facial key points to generate a position-enhanced feature sequence with superimposed embeddings; predicting the activation intensity and uncertainty of action units, and combining a static prior association matrix to perform causal intervention to generate a corrected feature sequence and a counterfactual feature sequence; and aggregating the corrected sequences to output the expression classification result. By dynamically adjusting the sampling distribution, the method focuses on high-discriminative regions and reduces noise; combines spatial structure and uncertainty estimation to suppress low-confidence features; and utilizes prior knowledge to perform causal inference to remove spurious correlations, verify causal sufficiency, and improve the model's generalization performance in complex scenarios.

This invention discloses a hallucination detection method combining semantic graph modeling and hierarchical uncertainty estimation, relating to the field of text hallucination detection technology. The invention aims to address the problems of inaccurate hallucination identification and low efficiency in existing text hallucination detection methods. The invention includes: acquiring text generated by a large language model and assigning hallucination risk labels to the text generated by the large language model; forming a training set by combining the text generated by the large language model and the hallucination risk labels; training a hallucination detection model using the training set to obtain a trained hallucination detection model; acquiring the text to be detected and inputting it into the trained hallucination detection model to obtain a hallucination risk score. This invention is used to detect hallucinations in text generated by a large language model.

A system and method includes determining uncertainty estimation in an object detection deep neural network (DNN) by retrieving a calibration dataset from a validation dataset that includes scores associated with all classes in an image, including a background (BG) class, determining background ground truth boxes in the calibration dataset by comparing ground truth boxes with detection boxes generated by the object detection DNN using an intersection over union (IoU) threshold, correcting for class imbalance between ground truth boxes and background ground truth boxes in a ground truth class by updating the ground truth class to include a number of background ground truth boxes based on a number of ground truth boxes in the ground truth class, estimating uncertainty of the object detection DNN based on the class imbalance correction, and updating output data sets of the object detection DNN based on the class imbalance correction.

The application provides a kind of based on multi-task learning's submarine pipeline flow field prediction method and system, comprising: using variable scaling physical information neural network framework, construct the multi-task learning model for solving cylinder flow problem;Total loss function of multi-task learning model is constructed, and the adaptive loss function weighting method based on uncertainty estimation is used and the weight growth factor is introduced, to dynamically allocate weights for each loss term in the total loss function, obtain the final total loss function;Physical information neural network is trained based on the final total loss function, and the trained physical information neural network is obtained;Based on the trained physical information neural network, the target submarine pipeline flow field is predicted, and the prediction result is obtained;Wherein, the prediction result includes the velocity component and pressure distribution of the flow field around the flow field.The application has higher prediction accuracy and reliability compared with the prior art when solving the problem of submarine pipeline hydrodynamic analysis.

A software tool for predicting a candidate recipe for a food product using: (a) a predictor model trained to output, for a given candidate recipe passed as input to the predictor model, (i) a predicted value for at least one target variable and (ii) a predicted value for a given subset of evaluation variables, and (b) a generator model for: (1) training a base prediction model configured to output (i) a predicted value for at least one target variable of a space of a possible recipe and (ii) an uncertainty estimate for the predicted value, and (2) selecting a candidate recipe from the space of possible recipes based on (i) a balance between a predicted value output by the base prediction model and the uncertainty estimate and (ii) a set of constraints.

This application provides a method, apparatus, device, and medium for autonomous driving data acquisition based on multimodal uncertainty fusion. The autonomous driving data acquisition method includes: acquiring multimodal data of the vehicle's surrounding environment in real time; inputting the multimodal data into an uncertainty estimation model, performing uncertainty fusion estimation processing on the multimodal data to determine a fusion uncertainty value; inputting the fusion uncertainty value into a reinforcement learning agent, generating a data acquisition optimization strategy based on the current environment state vector and the fusion uncertainty value; acquiring new data according to the data acquisition optimization strategy, and using the new data to adjust the autonomous driving model online. Based on the fusion uncertainty value predicted by the model, a data acquisition optimization strategy is dynamically generated to optimize autonomous driving data acquisition, improve data quality, and enhance system robustness.

This invention discloses an intelligent imputation method and system for continuous missing values ​​in time series based on context-aware generative adversarial networks. The method includes: preprocessing the original time series to generate a mask matrix; constructing a multi-scale dynamic contextual cue generator, extracting multi-scale features through a parallel temporal convolutional network, and dynamically fusing contextual information using a hierarchical attention mechanism to generate an adaptive cue matrix; constructing a time-aware generator, fusing the missing sequence, mask, noise, and cue matrix to generate imputed values ​​and uncertainty estimates; constructing a multi-scale discriminator; jointly training multiple relevant datasets using a multi-task learning framework, and jointly optimizing the model through adversarial, reconstruction, KL divergence, uncertainty calibration, and multi-task consistency loss; and using the trained model to impute missing data, outputting the complete sequence and the uncertainty at each position. This invention significantly improves the accuracy and generalization ability of continuous missing value imputation and provides reliable confidence assessment.

The application relates to a postoperative embolism syndrome monitoring method and system for a liver cancer patient, which comprises the following steps: acquiring multi-dimensional time sequence clinical data of the patient, interpolating missing values of the data through a probability regression model and generating uncertainty estimation for each interpolation point, constructing derived features based on the interpolated data, namely, calculating a time sequence change rate gradient of at least one core biochemical index and a time attenuation weighted fluctuation value of at least one vital sign index by taking the reciprocal of the uncertainty estimation as a weight, training a tree structure ensemble learning early warning model by using an asymmetric cost-sensitive loss function which is fused with a category imbalance penalty factor and a sample-specific weight determined by the time sequence change rate gradient of positive samples, inputting patient data to be warned into the trained model to output a risk probability, and triggering a high-level risk warning when the risk probability exceeds a preset threshold and key contribution features in model explainability analysis contain core inflammatory indicators as a subset of core biochemical indicators.

The present application relates to the technical field of deep learning, in particular to a Parkinson's disease prediction uncertainty estimation method and system based on model calibration, the method comprising: collecting multiple types of speech data from a subject sample and preprocessing to obtain traditional acoustic features corresponding to each type; inputting the traditional acoustic features into a prediction model for single-type alternating optimization to complete the optimization of the prediction model; obtaining the initial prediction confidence of the traditional acoustic features based on the optimized prediction model, and performing temperature scaling to determine the final prediction confidence of each type; determining the uncertainty of each type through the prediction confidence, analyzing the uncertainty to obtain integrated prediction results and uncertainty estimation, and feeding back the Parkinson's disease prediction of the subject sample; establishing a joint optimization target from the difficulty and mixed degree according to the prediction confidence, calibrating the prediction model through the joint optimization target, quantifying the prediction error of the confidence, and obtaining a classification framework with standard characteristics.

Systems and methods are directed to estimating depth uncertainty at a well location. The system can receive mistie data from a plurality of wells in a geographic region and determine a plurality of factors impacting depth uncertainty for each of the plurality of wells based on the mistie data. One or more machine learning models can be trained with the plurality of factors to calculate depth uncertainty for each of the plurality of wells. For a new well in the geographic region, the system can apply the trained one or more machine learning models to generate one or more graphical representations of depth uncertainty for the new well.

This invention discloses a method and system for predicting popular tourist cities, belonging to the fields of computer and tourism big data analysis technology. The method collects multi-source data, processes the data to construct multimodal features, constructs a weighted directed graph with cities as nodes, and performs multi-level aggregation of city node features to obtain a city representation that integrates its own features and information from neighboring cities. Based on this, a multi-task prediction network is used to predict the city's tourism popularity index within a target time window, and / or determine whether a city will become a popular tourist city within that time window, and performs uncertainty estimation. This invention integrates multi-source data, considers multi-scale inter-city propagation effects, utilizes a hybrid attention mechanism to improve the interpretability of prediction results, and enhances prediction credibility through uncertainty estimation. It can more accurately and proactively identify potential popular tourist cities than existing technologies, providing technical support for related application scenarios.

This invention discloses an intelligent detection method and system for data consistency in distributed energy systems, belonging to the interdisciplinary fields of energy internet, artificial intelligence, and industrial big data. The method includes: preprocessing multi-source heterogeneous data streams from the distributed energy system; achieving spatiotemporal alignment with uncertainty perception to obtain unified time-section data with accompanying uncertainty estimation; calculating device self-consistency scores by calling a physical constraint autoencoder, calculating topology consistency scores by calling a physical information graph attention network, calculating system-level consistency probability using a weighted minimum absolute value robust state estimation based on sparse measurements; and fusing multi-source evidence using Dempster-Shafer evidence theory to output the global consistency state of the distributed energy system data. This invention achieves accurate and interpretable fault location through a multi-dimensional, physically information-driven verification and fusion mechanism.

This invention provides an uncertainty quantification method and system for image classification tasks. The method includes: inputting an image to be processed into a neural network; treating each pixel of the image as a feature unit through the neural network; extracting the feature value of each feature unit to obtain an intermediate layer feature tensor of the image; introducing an implicit attention mechanism to determine the implicit attention potential of each feature unit; calculating a dropout probability matrix based on the implicit attention potential; discarding pixels of the image to be processed according to the dropout probability in the dropout probability matrix to obtain an image after dropping pixels; inputting the image after dropping pixels into a classifier; outputting the classification result of the image to be processed; and outputting a confidence evaluation index value characterizing the uncertainty of the classification result. Thus, efficient uncertainty estimation for image classification is achieved without increasing the number of model parameters or significantly increasing inference time.

A rotary kiln head temperature prediction method based on deep learning relates to the technical field of cement production preparation. It includes the following steps: determining the input and output of the temperature prediction model; preprocessing the original data, dividing the training set, validation set and test set by proportion; training the temperature prediction model through the training set, and adjusting and optimizing the hyperparameters in the model by using the MMGO meta-heuristic optimization algorithm; calculating the root mean square error MAE of the validation set on the temperature prediction model, dynamically allocating the weights of the CNN-BiLSTM module and the RF module in the whole model; using the Gaussian distribution method to calculate the probability interval, obtaining the probability proportion of the prediction value within a certain range, and then realizing the prediction uncertainty estimation; the trained temperature prediction model is used to predict in the test set. Advantage: it realizes higher precision, stronger generalization ability and uncertainty quantification of real-time prediction of rotary kiln head temperature.

This invention discloses a deep learning-based method for behavioral analysis and risk warning in clinical nursing processes, relating to the field of medical auxiliary diagnostic technology. It includes: S1: Constructing a three-branch deep neural network model through an individualized normal modeling path, a causal risk perception path, and a behavioral variation analysis path; S2: Setting orthogonality constraints and an uncertainty calibration assessment mechanism to obtain corresponding individualized normal vectors, risk feature vectors, style variation vectors, and uncertainty estimates, and combining them to determine the corresponding digital features; S3: Determining the corresponding anomaly confidence level and risk confidence level through the digital features, setting corresponding risk scores, and determining the corresponding risk level based on the risk scores and uncertainty estimates. This invention can effectively assist in the discovery of potential risks and ensure patient safety.

The application discloses a kind of medicine chemical reaction synthesis and conversion rate prediction combined optimization method, comprising: obtaining the SMILES expression of reactant, the SMILES expression of reactant is tokenized, and the tokenization feature of the SMILES expression of reactant is obtained by embedding expression;The tokenization feature of the SMILES expression of reactant is hierarchically sequentially encoded;Chemical reaction synthesis prediction and conversion rate prediction two tasks are combined, and two tasks are trained simultaneously to realize chemical reaction synthesis and conversion rate prediction combined optimization.The application combines chemical reaction synthesis prediction and conversion rate prediction two tasks, introduces hierarchical sequence modeling technology, and the model interpretability parameter optimization of two tasks guides each other, to improve the training efficiency and performance of model.The application introduces uncertainty estimation to cope with the interference brought by model to uncertain data in real situation.

The application provides a kind of based on small sample's air floatation synergistic oil-water separation effect prediction method and model, obtains the multiple original features of inlet water;According to the original feature generation expansion feature vector, the expansion feature vector at least includes original feature vector and physical interaction feature vector;Build base model, and the expansion feature vector is as the input of the base model;Adaptive optimization algorithm is used to train the base model;Through the prediction result of aggregation multiple base models, obtain the final prediction value and uncertainty estimation.The application still greatly enhances the ability of model to capture complex nonlinear relationship and prediction accuracy on the basis of small sample through polynomial expansion, physical interaction feature construction and deep neural network, effectively controls the model complexity by combining elastic network regularization and weight constraint, prevents overfitting, ensures that the prediction result conforms to the physical and chemical law through various constraints, improves the reliability and credibility of the model.

The application relates to the technical field of artificial intelligence, and discloses a migration-enhanced small-sample prediction method and system. The method realizes efficient knowledge migration from historical tasks to new tasks by constructing a memory-guided Bayesian meta-learning framework. The system comprises: a hierarchical memory bank for storing and organizing multi-level knowledge of historical tasks; a task representation learner for mapping different tasks to a unified embedding space; a memory retrieval and integration module for extracting relevant historical knowledge based on task embedding similarity; a Bayesian prior dynamic construction mechanism for converting the retrieved memory knowledge into a prior distribution suitable for the new task; a memory-guided Bayesian posterior inference module for integrating prior knowledge and new task data; and an uncertainty resolver for distinguishing cognitive uncertainty and data uncertainty. The application significantly reduces the amount of data required for the model to adapt to new tasks, improves prediction accuracy, provides reliable uncertainty estimation, and has self-adaptive migration capability, and is particularly suitable for data-limited social governance and other prediction scenarios.

The application discloses a dynamic working condition precise control method for gallium enrichment and impurity removal process, which comprises the following steps: firstly, an initial Gaussian prediction model is constructed based on historical working condition data, which is used for model predictive control and working condition monitoring; when the working condition is monitored to be switched, data is collected for model updating; in the transition stage, model mismatch causes conceptual drift of the prediction result, and the model predictive control performance is reduced; in order to reduce the control fluctuation in the transition period, the input of the regulation and control variable is tightly constrained according to the prediction uncertainty estimation; after a small amount of samples are collected, the model updating module obtains a drift matrix through a conceptual drift correction method, and generates a new working condition data set with pseudo labels in combination with the original sample set; subsequently, the model is reconstructed based on the new data set, and the working condition is monitored again; the model predictive controller adaptively relaxes the input constraint boundary of the regulation and control variable, and ensures precise control. The application can ensure the stability and control precision of the gallium enrichment and impurity removal process under the condition that the working condition frequently changes.