19 results about "Learning set" patented technology

The embodiment of the invention provides a federated learning contribution evaluation method and device, and the method comprises the steps: carrying out the grouping of a plurality of edge computing devices, and obtaining a plurality of sub-federated learning sets; and for the target federated learning sub-set, aggregating model update information corresponding to each edge computing device in the target federated learning sub-set, and determining a collaborative contribution value of the target federated learning sub-set based on the performance index of the updated global model on the common test set. Through a first linear programming solver, according to the collaborative contribution values of the multiple federated learning sub-sets, obtaining the maximum loss value corresponding to all the federated learning sub-sets and optimizing the maximum loss value to obtain the minimized maximum loss value, and through a second linear programming solver, obtaining the maximum loss value corresponding to all the federated learning sub-sets; and according to the maximum loss value after all the sub federated learning sets are minimized and the reference contribution values corresponding to the plurality of edge computing devices, target contribution vectors corresponding to the plurality of edge computing devices are determined, and the contribution degree of each edge computing device in the training process is accurately quantified.

This set provides an efficient scale sight-reading learning tool for piano beginners, allowing them to easily learn how to read musical scales on the staff and the correct key placement, even without a piano. [Solution] The scale reading learning set 1, which has a main body with four octaves of musical staff lines, is equipped with rotatably supported sheet winding knobs 4 and 5, and a transparent sheet for writing musical notes 6 that is stretched over the sheet winding knobs 4 and 5 via a transparent sheet for writing musical notes retraction hole 7. By rotating the sheet winding knobs 4 and 5, the musical note marks written on the transparent sheet for writing musical notes 6 move along the musical staff lines. This makes the scale reading learning set 1 compact, limits the range of the musical staff lines to four octaves, and improves the efficiency of learning to read musical notes by allowing the transparent sheet for writing musical notes 6 to move smoothly from side to side and allowing the written musical notes to be read immediately.

The application discloses a method for expanding random telegraph noise (RTN) based on a storage neural network, and the signal expansion process is realized based on an artificial neural network of a novel storage unit. According to partial RTN measured signals as a learning set, the expansion process of signal prediction reasoning can realize expansion of the signals with an arbitrary time length, and accelerates extraction of the time parameters of the RTN. The method has important significance for development of a physical unclonable function (PUF) technology based on the RTN and information data security.

The invention relates to the technical field of intelligent operation and maintenance of a power system, in particular to an extra-high voltage transformer substation handover test data fusion and abnormity early warning method, which comprises the following steps: collecting and preprocessing multi-source heterogeneous original data to generate a standardized test data set; based on a pre-trained multi-dimensional association rule model, hidden association rules among the test parameters are mined, weighted fusion is executed, and fusion data representing the overall health state of the equipment are generated; comparing the fusion data with a dynamic threshold interval generated by a historical normal sample, calculating a deviation degree and mapping the deviation degree into an abnormal confidence degree; and triggering visual early warning signals of different levels according to the abnormal confidence coefficient, and feeding back a new sample to the learning set for iteratively optimizing the multi-dimensional association rule model and the dynamic threshold interval. According to the method, deep fusion analysis of test data, dynamic quantitative evaluation of the health state and continuous optimization of the model can be realized, and the accuracy of anomaly recognition and the adaptability of the system are improved.

This application provides a federated learning contribution evaluation method and device. Multiple edge computing devices are grouped to obtain multiple sub-federated learning sets. For a target sub-federated learning set, model update information corresponding to each edge computing device in the target sub-federated learning set is aggregated. Based on the performance metrics of the updated global model on a public test set, the collaborative contribution value of the target sub-federated learning set is determined. Using a first linear programming solver, the maximum loss value corresponding to all sub-federated learning sets is obtained based on the collaborative contribution value of the multiple sub-federated learning sets, and the maximum loss value is optimized to obtain the minimized maximum loss value. Using a second linear programming solver, based on the minimized maximum loss value of all sub-federated learning sets and the reference contribution value corresponding to the multiple edge computing devices, the target contribution vector corresponding to each of the multiple edge computing devices is determined, accurately quantifying the contribution of each edge computing device in the training process.

The furnace temperature field reconstruction method based on the radiation transfer equation and the multi-layer feedforward neural network fusion comprises the following steps: calibrating the radiation detector by using a blackbody furnace to establish the relationship between the image intensity and the radiation intensity; establishing a detailed radiation transfer equation RTE considering the medium absorption, scattering and wall effect by using a DOM algorithm; generating a data set as a learning set of the neural network by smoothing the temperature field and the fluent simulation; constructing a multi-layer feedforward neural network; predicting the temperature of each grid by the learned model to obtain the furnace cross-section temperature field. The method couples the physical constraint of the radiation transfer equation with the feedforward neural network, avoids the shortcomings of the large amount of calculation of the pure physical constraint and the lack of physical theory support of the pure data driving, has small workload for replacing the fluent model simulation learning set for different sites, and can quickly obtain the temperature measurement result by calling the pre-learned model, and occupies less computing resources.

Various embodiments of the present disclosure provide machine learning architectures and optimization techniques for improving predictive functionality of a computer. The techniques comprise generating a cohort-level optimization dataset with a plurality of entity-level predictive features and a plurality of feature-level predictive features for a plurality of entity data objects using a machine learning ensemble model. The techniques comprise identifying a plurality of iterative candidate outputs through a series of optimization iterations. During each optimization iteration, an iterative candidate output may be generated by applying optimization model and a constraint set combination to the cohort-level optimization dataset. The techniques comprise selecting a target output from the plurality of iterative candidate outputs based on selection criteria.

Provided are a learning device that efficiently generates image data while suppressing an operation amount when an image is generated using a machine learning model using a variational autoencoder (VAE), and an inference device that executes a predetermined inference process on target image data using the machine learning model.SOLUTION: An image processing device that exhibits a function as at least one of a learning device and an inference device includes a first machine learning model to which input image data is input from an image input unit, a second machine learning model to which a latent variable generated by the first machine learning model is input and from which output image data is output, and a learning process execution unit that executes a process of learning a setting value with the two machine learning models. The execution unit converts input image data into a latent variable using a first machine learning model, generates output image data from the latent variable using a second machine learning model to learn a setting value, maintains a brightness component of at least one of the input image data and the output image data, and downsamples only a tint component.SELECTED DRAWING: Figure 2

A self-learning cluster control system for supply chain computing infrastructures, comprising a supply chain data acquisition unit, a node profiling unit, a cluster formation unit, a condition assessment unit, a self-learning control unit, and a resource allocation unit, wherein the system is configured to dynamically group supply chain nodes into clusters based on continuously acquired condition data and to enable adjustment of cluster allocation based on demand trends, inventory levels, transportation capacities, delivery times, and disruption risks.

This invention discloses a joint optimization method for edge computing-based vehicular network (V2N) resource management based on the DDPG algorithm. The method involves: establishing a network architecture integrating edge services and federated learning in the V2N, initializing network parameters; training a model using a dataset and calculating optimal model parameters; modeling the joint optimization problem as a Markov decision problem and training it using the DDPG algorithm, updating network parameters, recording the average reward, obtaining a decision network, and performing dynamic network offloading scheduling, resource allocation, and service model caching to achieve joint optimization of edge computing-based V2N resource management. This invention improves the real-time performance of V2N services based on edge computing and federated learning, exhibiting good convergence performance and joint optimization effects; it also enhances the security of privacy data on edge servers and the service quality for V2N users, and can be widely applied to practical mobile terminal applications such as path planning and navigation, and remote vehicle diagnostics.

The application provides a phenotype prediction method and system based on feature reduction and generalized inverse technology, which adaptively fuses linear and nonlinear effects, constructs a new model of adaptive fusion of linear and nonlinear effects of prediction factors for crop phenotype prediction, and evaluates the importance of different effects on the phenotype through the method of adaptive weight adjustment, so as to realize the function of comprehensively considering the influence of linear main effect and nonlinear relationship in crop phenotype prediction. When solving the model, the feature reduction is performed on the training data set, and the model is quickly and effectively solved by using the generalized inverse, so that the determination efficiency of the phenotype value of biological materials is improved. The learning set and the test set obtained by randomly dividing the experimental data set for multiple times are used for model learning, so that the system error of the model is reduced, and the stability of the phenotype prediction result is improved. The effectiveness of the application for phenotype prediction is verified based on DNA molecular marker genotype data and metabolite-based intermediate omics data.

The invention discloses a digital employee automatic construction method based on business system self-learning, and the method comprises the steps: receiving a natural language demand description of a business system through a self-learning integrated agent, calling a large language model to carry out intention understanding, and generating a structured tool development plan; the self-learning integrated agent adopts a three-stage self-learning strategy of exploration-learning-optimization to generate a parameterized script; the self-learning integrated agent encapsulates the optimized parameterized script into a standardized tool conforming to the MCP protocol specification, and encapsulates the structured business data model into a data structure definition document of the tool; the self-learning integrated agent registers the standardized tool to an MCP tool registration center; a digital employee agent loads a tool and business data structure definition from an MCP tool registration center, intelligent tool calling is carried out by adopting an'intention-planning-execution 'three-layer decision-making architecture, business task execution is completed, and the construction threshold and cost of digital employees are greatly reduced.

To provide a teaching material for facilitating composition in a language to be learned.SOLUTION: To provide a language teaching material set which enables a user to easily learn grammar by using composition cards on which composition terms of a learning object language and headings of speaker language concepts as superordinate concepts of the composition terms are described and arranging the composition cards corresponding to the headings of the speaker language concepts arranged in the word order of the learning object language on a composition sheet.SELECTED DRAWING: Figure 9

The present invention relates to a process abnormal state detection system of an artificial intelligence-based data integration structure. The process abnormal state system of an artificial intelligence-based data integration structure comprises: an image data acquisition module for acquiring image data; a sensing data acquisition module for acquiring process sensing data according to process parameters generated during equipment processing; an ensemble data model generation module for generating an artificial intelligence learning ensemble data model from the image data and the process sensing data; and an artificial intelligence-based process detection module for monitoring a process state on the basis of the artificial intelligence learning ensemble data model.

Systems, methods, and apparatuses for blocking subscription transactions for a given subscription from being applied to an electronic payment method, while allowing other transactions to be applied to the electronic payment method. Aspects further comprise training a machine learning model, based on historical transaction data, to predict subscription transactions, and updating the machine learning model based on incoming transactions. Aspects further provide for allowing a user to indicate they wish to cancel a subscription, blocking charges to the subscription while communicating with the merchant to cancel the subscription, and then removing the block after the subscription is confirmed canceled. to Aspects further provide for detecting when a subscription transaction was not blocked properly and updating the machine learning model to block similar subscription transactions in the future.

The present application relates to a method and device for guiding multi-round human-computer interaction tasks. The method receives real-time feedback data when a target user performs a human-computer interaction task, inputs a pre-trained dialogue generation model to obtain a target output dialogue, and outputs the target output dialogue. The model dynamically manages the training process by maintaining three data structures, namely a sample pool, a learned set, and a review buffer. Each iteration filters a training subset according to the sample difficulty evaluated by the current model, updates the gradient based on the historical samples, and dynamically classifies the samples according to the sample difficulty re-evaluated after training. The embodiments of the present application improve the convergence efficiency and generalization ability of the model, alleviate catastrophic forgetting, reduce the demand for training samples, and improve the guiding effect and task completion rate of multi-round human-computer interaction.

The application discloses a GFM / GFL hybrid control collaborative optimization system and method for weak grid high-penetration new energy grid connection, relates to the technical field of new energy grid connection control, and a state acquisition and system strength evaluation module is used for collecting operation state data of a grid connection point in real time, estimating background impedance and calculating system strength; a two-stage distribution robust scheduling module in an upper layer performs multi-objective scheduling, and first-stage decision variables are obtained; a middle-layer safety reinforcement learning setting module generates proportion correction and control parameter actions, and real-time proportion coefficients are obtained; a master-slave hybrid control execution module executes master-slave GFM / GFL hybrid control based on the real-time proportion coefficients, a GFM channel provides a reference voltage reference, a GFL channel generates a correction current and maps the correction current into a correction voltage through a virtual impedance; and a safety guardian and rollback module switches system control modes. The application simultaneously considers economy, stability and power quality, and improves controllability and engineering deployment reliability under an extreme weak grid working condition.

The invention relates to an aerodynamic characteristic six-component modeling method based on a functional learning method, which belongs to the technical field of aircraft aerodynamic characteristic modeling, and comprises the following steps: firstly, establishing an initial data set, performing data identification and screening, and removing unreasonable data; performing mathematical transformation, and dividing the data into a learning set and a verification set; constructing an initial model according to the data in the learning set, and performing verification through a verification set; then evaluating the model; and finally, carrying out lightweight processing on the verified model, and finally outputting an aerodynamic coefficient display function model. Cross-wind-tunnel and cross-scale data training is fused, the model can break through limitation of a single wind tunnel flow parameter coverage range, six-component prediction under the conditions of a wide-domain Mach number and an attack angle is achieved, and generalization errors are reduced compared with a traditional empirical model. According to the method, the dependence on mass data is remarkably reduced through feature parameter dimensionality reduction and function space constraint on the basis of a physically guided symbol regression framework, and the efficiency is greatly improved by relying on analytic model characteristics.

The embodiment of the invention discloses an unsupervised mixed domain separation method and device for aliasing seismic data, and the method comprises the steps: carrying out the random time delay coding of original common-shot-domain multi-seismic-source aliasing data according to a seismic channel, and obtaining the randomly delayed common-shot-domain multi-seismic-source aliasing data; constructing a common-shot-domain unsupervised learning set by using the randomly delayed common-shot-domain multi-seismic-source aliasing data and the original common-shot-domain multi-seismic-source aliasing data; constructing a common detection domain secondary random aliasing unsupervised learning set by using the randomly delayed common shot domain multi-source aliasing data and the original common detection domain multi-source aliasing data; performing self-supervised cross iteration training by using the two learning sets to obtain a trained aliasing seismic data separation model; and inputting common-detection-domain multi-seismic-source aliasing data to be separated into the model for separation processing to obtain result data after aliasing separation. According to the invention, an unsupervised algorithm is adopted, and unsupervised intelligent processing of aliasing seismic data is realized.