49 results about "Model set" patented technology

The application discloses a three-dimensional model fast retrieval method, equipment and medium, and relates to the computer technical field. The method comprises the following steps: in response to receiving a part of a two-dimensional sketch of a three-dimensional model drawn by a user, performing image segmentation on the part of the two-dimensional sketch to obtain a plurality of sub-images; performing mapping processing on each sub-image to obtain an initial feature corresponding to each sub-image, and performing attention calculation on the initial features corresponding to the plurality of sub-images to obtain a plurality of first key features; fusing the plurality of first key features to obtain a second key feature corresponding to the part of the two-dimensional sketch; based on the second key feature, performing similarity matching on a feature database to obtain a plurality of target features, and searching a three-dimensional model set corresponding to the plurality of target features from a three-dimensional model database. In this way, the target three-dimensional model meeting the design intention of the user can be quickly and accurately retrieved according to the part of the sketch of the model drawn by the user, so that the design efficiency and the user experience are significantly improved.

The application discloses a malicious code model detection method and device and a computer readable medium. Local models generated by respective local model training of each edge side device are obtained to obtain a local model set. Similarities between a single local model in the local model set and other local models except the single local model are determined to obtain a similarity set corresponding to each single local model. According to the similarity set corresponding to each single local model, a local model with a similarity satisfying a preset condition is determined from the local model set as a malicious code model. The application solves the risk diffusion problem caused by the malicious code model by calculating the similarities between the local models and detecting the local model with the similarity satisfying the preset condition as the malicious code model.

The present application relates to the field of shell-and-tube heat exchanger, in particular to a kind of multi-physical field coupling performance comprehensive simulation evaluation system of shell-and-tube heat exchanger, system includes: analytical modeling module;For integrating obtaining semantic parameterization geometric information library;Flow field geometric model repair and parameterization construction module;For completing the topological reconstruction of initial flow field geometric model and obtaining zero defect parameterization three-dimensional flow field model;Cross-domain grid collaborative generation module;For integrating forming cross-grid model set;Multi-physical field simulation solving and evaluation module;For forming multi-physical field simulation original data set, simulation evaluation is carried out.The present application will flow field geometric model repair and parameterization construction module be converted into three-dimensional voxel grid by voxelization discrete to model, avoid model geometric feature distortion by fairing processing etc., solve the problem that traditional model repair means is single, topological distortion, defect identification is not accurate, guarantee the geometric accuracy of subsequent simulation calculation from source.

Feature importance is critical to understanding how predictive models produce accurate results, and can change significantly for different models. The present invention is used to achieve a good ranking for stable feature importance. An optimized technique is presented which considers feature importance value variation within different groups of cross-trained models. Feature importance is computed for all group models with this optimized method, and then a best set of models can be selected based on classification error as well as optimized stable feature importance values.

The application discloses a modeling method and device of equipment, a processor and an electronic device. The method comprises the following steps: determining a plurality of components of a target equipment, wherein the target equipment is an equipment to be modeled in three dimensions, and the components are at least one of the following: components and boards; acquiring a component 3D model set and a board 3D model set, wherein the component 3D model set at least comprises 3D models of a plurality of components, and the board 3D model set at least comprises 3D models of a plurality of boards; determining a 3D model of each component based on the component 3D model set and the board 3D model set; and obtaining a 3D model of the target equipment according to the 3D model of each component. Through the application, the problem that the effect of 3D modeling of equipment is poor in the prior art is solved.

The application discloses a set smoothing time-lapse seismic difference inversion method based on probability mode conversion. The method comprises the following steps: acquiring time-lapse seismic data and logging data; constructing a linear forward model between difference seismic data and relative wave impedance difference parameters based on a convolution model; extracting a logging label reflecting real underground wave impedance difference, calculating a cumulative probability distribution and a functional inverse function thereof based on a probability integral transformation, and constructing a probability mode conversion operator; in a set smoothing framework, updating a model set of a multivariate Gaussian distribution by using a Kalman gain, and mapping the updated model to a model set conforming to a logging real distribution through the conversion operator; and iteratively performing the above process to obtain a posterior inversion result of the relative wave impedance difference. The application can directly invert a relative difference parameter, and effectively improves the inversion precision of a time-lapse seismic difference parameter with a non-Gaussian distribution characteristic and the reliability of reservoir dynamic characterization.

The embodiment of the application provides a cluster model calling path generation method and product based on a network body intelligent architecture, and relates to the technical field of artificial intelligence. On the basis of obtaining model multidimensional data of each cluster model in a model cluster, model multidimensional features corresponding to each cluster model are generated according to the model multidimensional data of each cluster model, and model correlation features between the cluster models are generated. Model path features between the cluster models are generated according to historical running path features, the model correlation features between the cluster models and the model multidimensional features corresponding to each cluster model. The model multidimensional features of each cluster model and the model path features between the cluster models are summarized to obtain multidimensional features of the model cluster. A model calling path is generated according to a user task request and the multidimensional features of the model cluster. That is, accurate matching of the model calling path and efficient scheduling of the cluster model are realized, and the problem of poor matching of the calling path is solved.

The application discloses a kind of quick and accurate detection methods of sheltered apples, the method is based on deep learning instance segmentation model, and single apple is segmented from RGB image, and during model training, by increasing model branch and artificial labeling, guide model to learn how to extract, fuse and comprehensively consider the features of the apple around target apple, improve the identification ability of model to sheltered apple.The application effectively reduces the decline of detection model on the detection performance of dense, mutually sheltered apples, and the performance decline on the public dataset is only 50% of the original, improves the overall performance and robustness of the algorithm, can make accurate segmentation detection on sheltered apple, and the algorithm runs fast, only needs a little extra calculation, can be easily integrated with other models.

The application discloses a method for predicting performance of repair materials and optimizing formula based on a machine learning algorithm, and particularly relates to the field of cross between artificial intelligence and material science, comprising the following steps: constructing a structured multi-modal feature library, adopting double indexing and differential noise filtering for data processing, constructing a heterogeneous model set guided by physical mechanism, including three targeted sub-models of a graph neural network, a time series convolution network and a gradient boosting tree, dynamically fusing outputs of the sub-models through a meta-learner to realize joint prediction of material performance, actively searching for potential formula in a high-dimensional solution space through dimension reduction mapping and Bayesian optimization, and generating an optimal formula scheme by using a differentiable physical-data fusion simulator for gradient-guided constraint optimization, and finally realizing automatic feedback of new data and self-evolution of the model through triggering rules and local incremental learning mechanisms based on uncertainty and deviation, so as to form a complete self-evolution closed loop system.

The application provides a model setting method and system of a free report, and a readable storage medium. The model setting method of the free report comprises the following steps: acquiring a cell list, the cell list comprising a plurality of design cells; converting the design cells to obtain business cells; structurally processing the business cells to obtain design blocks; and sequentially processing cell information in the design blocks, and collecting cells on the same link together according to the cell information to form a design path. Through the technical scheme, the server adds the design block layer and the design path layer in the running time, the execution is clear in hierarchy, supports a more fine-grained business scenario, and is convenient for subsequent code maintenance and expansion.

The application discloses a model checking quantity determination method, device and equipment, and the method comprises the following steps: obtaining model information and checking parameters corresponding to a plurality of three-dimensional models respectively; classifying the plurality of three-dimensional models according to the model information, and determining a plurality of model sets; determining the similarity and complexity between the plurality of three-dimensional models according to the model information; and determining the checking quantity for indicating the quantity of to-be-checked models in each model set according to the checking parameters, the similarity and the complexity. The application adopts unified checking parameters to determine the checking quantity of different types of model sets, adopts a unified sampling standard to determine the checking quantity of each type of model, avoids the waste of operation resources and time cost caused by manual participation, and improves the accuracy of model sampling.

The present application describes a method to determine an architecture for multi-task learning based on feature similarities. The proposed method is characterized by comprising the steps of: training each model of the set of models in a single-task; perform a first evaluation of feature similarities between each single-task trained model; training each model of the set of single-task trained models in cross-tasks; perform a second evaluation of feature similarity between the set of cross-task trained models; output a multi-task learning architecture model based on feature similarities from the cross-task trained models.

The application provides an AI Agent-based text-to-image flow process decision execution method and device and a storage medium. The method comprises the following steps: analyzing a workflow file of a text-to-image tool, extracting a node set, a data link connection relationship between nodes, a model set relied on by the nodes, and a dependency relationship between the nodes and the models; comparing the node set and the model set with installed nodes and deployed models of a target platform to determine a fault type; determining a fault handling sequence by using an AI Agent with a built-in knowledge base module; performing a self-healing operation by using the AI Agent according to the fault handling sequence; starting a trial operation verification on the workflow after the self-healing operation is completed, and updating the fault type, the handling sequence and the self-healing operation to the knowledge base module. By using the application, cross-environment migration of the text-to-image workflow is achieved, and the migration efficiency and success rate are improved.

This invention discloses an online prediction method for battery SOC based on an incremental correlation vector machine (RVM) strategy using a multi-kernel ensemble approach. The method includes the following steps: Step 1, data preprocessing; Step 2, training set sampling; Step 3, kernel function selection; Step 4, model training; Step 5, model validation; Step 6, adaptive kernel parameters; Step 7, RVM model ensemble; Step 8, model prediction; Step 9, incremental learning strategy; and Step 10, online incremental prediction. From a practical perspective, this invention addresses the complexities and diverse needs of various applications. Drawing on the ideas of incremental learning and ensemble learning, it generates highly differentiated RVM individual learning models containing multiple kernel functions through dual perturbation of training samples and kernel functions. Combined with a novel incremental ensemble strategy, it avoids the problem of model overlearning, improves the model's generalization ability and robustness, and expands its application scope.

This invention discloses a defective code block localization method based on model structure evolution. First, based on several deep learning computing framework libraries, its object libraries and security issues are summarized to construct a computing framework object library. Then, a computing framework security defect library and a code security defect library are constructed. Next, a neural network model is constructed and mutated to obtain a set of mutated models. Output differences are extracted from the mutated model set to obtain feature difference vectors, and the prediction matching rate is calculated to determine whether the neural network model contains defective code blocks. Finally, the defective code in the neural network model is verified and traced using statement manipulation methods, and compared with the computing framework security defect library and the code security defect library to achieve defective code block localization. This invention reduces computational costs, has wider deployment and application scenarios, and improves the accuracy of code block defect localization.

The application discloses a kind of academic question and answer-oriented multi-model paper retrieval method, it is related to natural language processing and information retrieval technical field, including: first, construct unified corpus and training dataset, utilize the model in first model set and second target model respectively encode generation document vector set;Then, based on the initial retrieval result of second model, difficult negative sample is filtered, and the contrast learning sample pair is constructed to fine-tune and re-encode corpus;With the model group of the second target model after fine-tuning and the model in first model set, each model in model group is executed similarity retrieval in parallel respectively, and the corresponding original similarity matrix is obtained, based on the original similarity matrix, the document is filtered, and the first target document list of target query is generated.The application can effectively improve the accuracy and robustness of academic literature retrieval.

The application provides a leaf area index inversion uncertainty quantification method and system, and belongs to the technical field of remote sensing data processing, and comprises the following steps: extracting target spectral features according to a remote sensing image; generating a simulation sample set based on a radiation transfer model; combining model sensitivity and parameter prior calculation to obtain a first uncertainty degree derived from physical parameter transmission; repeatedly training and constructing a machine learning model set by using the simulation sample set; calculating a second uncertainty degree derived from algorithm randomness according to the statistical dispersion degree of model set prediction values; and obtaining a total uncertainty degree according to the first and second uncertainty degrees. By quantifying physical modeling errors and algorithm randomness errors respectively, end-to-end decoupling and comprehensive of the uncertainty are realized, so that the accuracy of the confidence evaluation of the leaf area index inversion result is improved.

The present application relates to the cross field of cloud computing and artificial intelligence technology, in particular to a tool deployment method, system and device based on model context protocol, aiming to solve the problem of complex integration process of scientific research tools and large language models (LLM), the present application sets tool parameters and logic by using graphical interface, and realizes automatic generation of tool description file conforming to model context protocol specification; in order to meet the rapid and efficient deployment effect of scientific research tools, a hierarchical construction strategy is adopted, and a container image is generated according to the running environment dependency list in the tool description file, so as to improve the tool construction rate; at the same time, based on the security information field and resource requirement list in the tool description file, according to the preset allocation rule, the appropriate running instance is dynamically allocated to ensure the safety in the tool deployment process, and at the same time, the labor cost is reduced, and the integration, deployment and running efficiency of tools and large language models are improved.

This invention provides a question-and-answer data evaluation method, system, electronic device, and storage medium. The method includes: using any language model from a model set as an evaluation model; inputting evaluation instructions for the question-and-answer data to be evaluated generated by a first model into the evaluation model to obtain a preliminary evaluation result generated by the evaluation model; driving the evaluation model to perform multiple rounds of iterative correction based on at least one second model, starting from the preliminary evaluation result, until the evaluation model is corrected to obtain a target corrected evaluation result with converged evaluation scores; and combining the target corrected evaluation results obtained when each language model in the model set is used as the evaluation model to generate a quality evaluation report of the question-and-answer data to be evaluated. This invention significantly improves the objectivity, stability, and accuracy of question-and-answer data quality evaluation by introducing domain role alignment, cross-evaluation, and a reflective correction mechanism based on multi-round iterative questioning among multiple models.

This invention discloses a design method and apparatus for refractory high-entropy alloys based on a large model, relating to the field of alloy material design technology. The method includes: First, by generating and collecting data, pre-trained surrogate models for thermodynamics (covering phase diagrams, solidification processes, etc.) and mechanical properties are constructed, and calculation models for physical parameters such as valence electron concentration and density are established. Then, these models are integrated into a single ensemble model, and a large language model is introduced as the core of intelligent scheduling. The large model is responsible for parsing the design task, intelligently planning and calling the interfaces of each model based on a knowledge base and rule base. Finally, based on the set design objectives, a multi-objective genetic algorithm is initiated for iterative optimization and adaptive adjustment, efficiently outputting a final alloy design scheme that meets multiple performance requirements. This invention enables intelligent multi-objective design of refractory high-entropy alloys, solving the problem of high dependence on manual intervention.

The application is suitable for the field of automobile aerodynamic performance prediction technology, and provides a cross-dimension 2D-3D automobile aerodynamic prediction method based on an artificial intelligence algorithm. First, sensitive parameters of geometric features under the normal viewing angle of the automobile are selected for parameterized deformation to generate a parameterized deformation table and an automobile model set, and two-dimensional and three-dimensional CFD simulations are respectively carried out. Second, based on the mapping rule of the parameterized deformation table and the aerodynamic force coefficient, a K-Means and GMM hybrid clustering strategy is adopted for sample classification, the optimal clustering category number is determined in combination with MAE and MSE indexes, and a multi-branch neural network model set is constructed and trained for each type of sample. Finally, the deformation parameters of the new three-dimensional vehicle body model to be predicted are input into the model set, after clustering positioning and branch regression preliminary prediction, a scene-specific cross-dimension gain coefficient is introduced for correction and compensation, and the resistance coefficient prediction value is output. The method can greatly reduce the three-dimensional CFD simulation workload, ensure the prediction accuracy, and significantly save the computing resources.

The application discloses a kind of motorized target fusion tracking methods for full-dimension missing dimension disorder measurement, in view of the problem of disorder measurement data in update cycle, introduce non-sequential measurement one step lag filter method;In view of the characteristics that measurement data exists full-dimension, optimization filter update model, realize the coverage of each configuration sensor measurement data under unified framework;For the problem of complex target maneuver, build a smaller model set that can be applied to engineering, introduce interactive multiple model algorithm.Finally, realize the efficient use of heterogeneous sensor disorder measurement data, improve the tracking ability of motorized target.

The application discloses a subject state recognition method and device based on a multi-dimensional model fusion framework and electronic equipment, wherein the recognition method divides the data of a historical subject into multiple dimensional data sets; the data subsets in the data sets are combined in multiple dimensions to obtain multiple combination sets; the recognition models are trained according to the combination sets to obtain a model set containing multiple recognition models; the recognition models in the model set are fused to obtain a fusion recognition model; and the to-be-recognized subject is recognized through the fusion recognition model to determine the state of the to-be-recognized subject. According to the scheme, the recognition models of multiple dimensions are fused to obtain a fusion recognition model, the state of the to-be-recognized subject is determined through the fusion recognition model, the state of the to-be-recognized subject can be quickly and accurately supervised and evaluated, early warning information can be generated in time, and the information safety, data transmission safety and system safety of the subject and the service platform are ensured.

Logic may identify feature contributions to erroneous predictions by predictive models. Logic may provide a set of two or more models. Each model may train based on a training dataset and test based on a testing dataset and two or more models may be unique. Logic may test the set during a monitoring period. Logic may perform residual modeling on each model in the set during the monitoring period and may determine a list of input features that contribute to a residual of each model of the set. A residual comprises a difference between a predicted result and an expected result. Logic may generate a combined list of the input features from the set and may rank the input features. Logic may perform a voting process to generate the ranks for the input features. And logic may classify features as exogenous or endogenous based on a threshold and the ranks.

The application relates to the technical field of model construction, and particularly discloses a cardiovascular disease risk prediction model construction method based on a taper index, which comprises the following steps: S1, acquiring a sample data set, wherein each sample contains an outcome label and a feature vector, and the feature vector contains a continuous variable taper index; preprocessing the sample data; S2, dividing the preprocessed sample data set into a training set and a test set; constructing a candidate model library, combining the models in the candidate model library, and generating a candidate model set; S3, using a unified training interface to execute training on each model in the candidate model set, and extracting a feature subset reserved by each model after the training is completed; and S4, evaluating each trained candidate model on the test set, and screening an optimal cardiovascular disease risk prediction model according to the evaluation index result. The technical scheme of the application can solve the problems of non-systematic obesity and lipid index screening and insufficient generalization of a prediction model in existing cardiovascular disease risk prediction.

A sequential feedback prediction system predicts a risk of an event associated with a set of potential outcomes. The system receives training data including data records that have information associated with risk factors. Each data record may be labeled with outcomes that have inherent sequential dependencies. The system trains sub-models, each sub-model predicting risks for a respective outcome using a set of risk factors. For each sub-model, the system may determine a subset of sub-models whose results are used as input for the sub-model. The system determines an order to run the set of sub-models such that prediction results for preceding sub-models may be used as inputs for subsequent sub-models. The system may determine a number of rounds to run the ordered sequence of sub-models until the performance of the ensemble model meets a predetermined threshold. The system may generate a risk score for each of the set of outcomes.

The present application belongs to the technical field of cross of mine safety monitoring, rock mechanics and artificial intelligence, and particularly relates to a filling body stage intelligent identification method and system based on multi-model integration and transfer learning. First, the acoustic emission monitoring system of the filling body is built to collect the acoustic emission signals of the filling body in the whole loading process and extract multi-dimensional features; the improved differential evolution algorithm is used to optimize the piecewise linear regression model, and the four stages of filling body failure are automatically identified in combination with physical constraints to generate the source domain data set; then the key features are extracted through the anti-leakage feature optimization mechanism, and a soft voting heterogeneous ensemble model is constructed by using four kinds of heterogeneous base learners; aiming at the difference of data distribution of the target domain, the domain offset degree is calculated and the fine-tuning transfer strategy based on instance-level weighting is adopted to quickly update the model with a small amount of data. The present application solves the problems of strong subjectivity of artificial labeling, poor generalization of single model and difficult cross-domain adaptation, and realizes intelligent early warning of the whole process of filling body instability.

The application discloses a CALPHAD-based cross-system knowledge transfer alloy design method, which comprises the following steps: performing CALPHAD balance calculation on a source aluminum alloy and a target Al-Mg-Zn alloy to construct an initial feature set; deriving mean value and variance of phase fraction weighting and multi-step heat treatment addition / difference features; locking an optimal feature subset through three-step feature selection assisted by source data set; training 300 base models through bagging, integrating mean value and standard deviation to obtain a high-stability predictor; and coupling NSGA-II and high-throughput CALPHAD to search for a Pareto frontier with double targets of strength and elongation and determine a composition-process scheme; and the application significantly improves the model interpretability and generalization performance by constructing a CALPHAD descriptor with physical interpretability as a model feature, thereby improving the alloy design efficiency and reducing the experimental workload.