19 results about "Shared learning" patented technology

The present invention relates to the field of power automation. Disclosed are a power grid regulation and control model training method and system, a computer device, and a storage medium. The method comprises: acquiring training data of a power grid regulation and control model, and grouping the training data to obtain a plurality of training data groups; training a plurality of preset learning bodies in parallel by means of the plurality of training data groups on the basis of a one-to-one correspondence between the training data groups and the learning bodies to obtain a plurality of execution bodies, wherein when the learning bodies are trained by means of the training data groups, the training data groups are divided into a plurality of training data units, a training effect of each training data unit is acquired to obtain a shared training data unit, and the shared training data unit is shared with other training data groups than the training data group to which the shared training data unit belongs; and finally, combining the plurality of execution bodies to obtain a power grid regulation and control model. By means of parallel training and parallel prediction, quality is improved on the basis of quantity, and shared learning of high-quality sample data is achieved, thereby effectively improving a training effect, and being capable of effectively reducing the training time of the power grid regulation and control model under a big data condition.

The invention provides a substation terminal network equipment defect elimination checking method and system based on deep learning, and the method comprises the following steps: constructing an equipment configuration heterogeneous graph network by a substation client through combining an equipment configuration file, an equipment operation log and equipment network flow data, all the transformer substation clients utilize a heterogeneous graph network to train a local model, the federated learning server updates a shared learning model by combining local training parameters of all the clients, and the clients needing to be checked continue to update the model according to updating parameters of the server; and identifying a configuration mode of local network equipment through the updated model and obtaining a configuration parameter mapping table, generating an equipment checking strategy in combination with the configuration mode and the configuration parameter mapping table, and finishing checking work of all the network equipment through the equipment checking strategy. According to the invention, the transformer substation can accurately complete network equipment checking.

This invention discloses a heterogeneous government data collection and analysis system based on data governance and privacy protection, belonging to the field of intelligent analysis technology. The system includes: a data collection and cleaning module, which intelligently collects and cleans heterogeneous government data using adaptive algorithms; a privacy protection and dynamic desensitization module, which achieves comprehensive data privacy protection through dynamic desensitization technology and secure multi-party computation technology; an intelligent analysis and collaborative computing module, which, by introducing federated learning technology, enables multi-party collaborative model training without sharing original data, and shares the learned model parameters to achieve cross-departmental and cross-regional joint data analysis; and a data sharing and management module, which establishes a cross-departmental data sharing and verification mechanism through blockchain technology to ensure the transparency and traceability of data exchange. This invention can achieve efficient and intelligent government data collection, privacy protection, and analysis with low resource consumption.

The invention discloses an intelligent glaucoma diagnosis method based on multi-task learning. According to the method, a multi-task learning network model is constructed, and the model is composed of a U-Net network and a VGG network. The U-Net network is responsible for extracting features from the eye fundus image to obtain a cup-to-disk ratio, and inputting the cup-to-disk ratio as one of the features of the eye fundus image into the VGG network, and the VGG network performs glaucoma classification on the eye fundus image by using the features. By sharing the encoder portion of the U-Net network, the two networks can share the learned feature information. After being trained by a fundus image data set, the model obtains very high glaucoma classification accuracy. According to the invention, an accurate and efficient glaucoma diagnosis function is realized, the method can be widely applied to large-scale glaucoma screening tasks, the work of ophthalmologists is greatly facilitated, and the harm of glaucoma to people is reduced.

Provided is a federated learning method and system using shared learning data. The federated learning method includes determining shared learning data to be provided to federated learning devices, learning each of artificial intelligence (AI) models by each of the federated learning devices using the shared learning data and unique learning data of each of the federated learning devices, and determining a final model by a final model determination server based on a learning result of the AI models.

Provide a system. 【Solution means】 Means for collecting learning information of individual learners, Means for analyzing the collected learning information to identify the learners' strong and weak areas, Means for generating an individualized learning plan suitable for the learner based on the identified strong and weak areas, Means for presenting the generated individualized learning plan to the learner, Means for providing an immediate response to inquiries from the learner, Means for notifying the learner's learning progress, Means for providing an individually optimized learning experience in the local community where the learner resides, Means for sharing the learner's educational progress with parents or educational institutions, A system including the above.

The present disclosure provides a push method and system based on joint prediction, the method comprising: obtaining first data of a requesting party at a first participant and second data at a second participant, wherein the first data at the first participant is different from the second data at the second participant, and the first data and the second data are not interoperable; performing a joint prediction based on the first data and the second data obtained by the shared learning combination to perform joint prediction-related push to the requesting party; and feeding back the result of the joint prediction to the first participant and the second participant.

This invention discloses a method and apparatus for compressing and inheriting ancestor model knowledge into descendant models through optimal transmission. The method includes: constructing an ancestor model containing multiple Transformer network layers and obtaining a unified feature representation for each layer of the ancestor model; constructing a difference metric between each layer of the ancestor model and candidate learning genes; modeling the cross-layer knowledge mapping process as an entropy-regularized optimal transmission problem; solving the entropy-regularized optimal transmission problem to obtain a cross-layer knowledge transmission matrix; performing cross-layer weighted fusion of ancestor model parameters based on the cross-layer knowledge transmission matrix to construct shared learning gene parameters; constructing descendant models of different scales based on the learning gene parameters; and applying the constructed descendant models to the training and inference of downstream tasks. This invention can complete structured knowledge inheritance and model reconstruction without accessing the original training data of the ancestor model, and has good deployability and generalization ability.

The application discloses a kind of multilayer convolutional brain-like chips based on full pulse HMAX model, chip uses layer type configurable multi-core architecture, overall by global controller, several cascaded pulse convolution kernel, interlayer shared learning engine and output pulse decoder four parts are formed.The global controller is mainly responsible for the transmission, interaction and control of processor internal and external data.Pulse convolution processing kernel is the core module of processor, it is connected in turn, and the pulse data output by the previous processing kernel is received by the next processing kernel.When a certain network layer is in learning state, the network layer will call interlayer shared learning engine module to complete on-chip learning.Output pulse decoder decodes AER data output, thereby realizing classification function, and the result is output externally.The application can realize efficient on-chip learning and inference of full pulse HMAX model, and solves the problems of low recognition rate, insufficient performance and poor energy efficiency of existing edge brain-like chips.

The embodiment of the invention provides a river water quality detection method and system based on artificial intelligence, and the method comprises the steps: obtaining the field collection data of a river water quality detection system as regional water quality monitoring samples, and carrying out the water quality state label prediction of the regional water quality monitoring samples through a shared learning network; the accuracy and the efficiency of water quality detection are obviously improved. Specifically, a deep learning network with the best training effect is selected from a plurality of training entities participating in linkage network parameter learning as a shared learning network, so that the reliability and high precision of water quality state label prediction data are ensured. Furthermore, parameter learning is performed on the target water quality evaluation model according to the prediction data, and an optimized target application program is determined, so that more accurate water quality state label prediction data can be generated. According to the method, not only is the intelligent level of water quality detection improved, but also personal errors and time cost possibly existing in a traditional water quality detection method are effectively reduced.

The application discloses a loading and unloading equipment health state monitoring method based on multi-similar equipment shared learning, real-time collection of operation data of multiple similar loading and unloading equipment, server initialization of a basic model for each loading and unloading equipment, self-adaptive learning of each loading and unloading equipment according to collected data and performance evaluation of the basic model, server fusion of the basic models of the multiple loading and unloading equipment into a global model and generation of an individualized model according to local data of each loading and unloading equipment, health state prediction by using the individualized model and provision of early warning information, maintenance plans and suggestions according to the health state prediction result. The application fully utilizes the similarity between the loading and unloading equipment in structure and function, realizes deep mining and effective utilization of a large amount of monitoring data in a shared learning mode, and greatly improves the accuracy and real-time performance of monitoring.

The present invention provides a method and system for verifying network equipment at a substation end based on deep learning. The method comprises the following steps: a substation client constructs a device configuration heterogeneous graph network in combination with the device configuration file, the device operation log, and the device network traffic data; all substation clients use the heterogeneous graph network to train local models; a federated learning server updates a shared learning model in combination with the local training parameters of all clients; clients that need to be verified continue to update the model according to the updated parameters of the server, and identify the configuration mode of the local network device through the updated model and obtain a configuration parameter mapping table; a device verification strategy is generated in combination with the configuration mode and the configuration parameter mapping table; and verification of all network devices is completed through the device verification strategy. The present invention has the effect of enabling substations to accurately complete network equipment verification.

The invention provides a spatial multi-omics integration method and device, equipment and a storage medium. Relates to the technical field of spatial multi-omics. The method comprises the steps of obtaining a double-omics data set, and constructing an adjacent graph containing a space graph and a feature graph based on the double-omics data set; by means of an in-omics integration module, graph convolution is performed on the space graph and the feature graph, space and feature potential representations are spliced and input into a multi-layer perceptron, and in-omics fusion representation is obtained; through an inter-omics integration module, performing weighted summation on the intra-omics fusion representation by using a multi-head attention mechanism to obtain a multi-omics potential representation; a space omics integration model is trained by means of a double learning strategy, multi-omics potential representation and original normalized space mapping are guaranteed through private learning in omics, specific feature coding is enhanced by introducing self-supervised contrast learning, alignment of different omics representations is restrained through shared learning among omics, and space multi-omics integration is achieved. According to the method, more complex histological structure characteristics can be captured, and the analysis capability of a complex space structure is remarkably improved.

The subject disclosure relates to techniques for enabling sharing of knowledge among a fleet of autonomous vehicles. A process of the disclosed technology can include generating an update for a continuous deep learning neural network on-board the autonomous vehicle based on driving scenarios encountered by the autonomous vehicle during its deployment and providing the update for the continuous deep learning neural network to additional vehicles in the fleet on autonomous vehicles, wherein the update for the continuous deep learning neural network is configured to be incorporated into a joint kernel for use by the additional vehicles in the fleet on autonomous vehicles.

The invention relates to a data quantitative analysis and factor analysis method. According to the technical scheme, the method comprises the following steps: firstly, classifying original data, and identifying uniform distribution, skewed distribution and periodic distribution; according to the data distribution, generating a dynamic quantitative standard through a variational auto-encoder VAE model; quantizing the data according to the generated standard, so that each data point is mapped to a quantized value conforming to the feature; taking the quantized data as input; establishing a model for capturing a nonlinear factor relationship in data by introducing a deep neural network; extracting potential factors of the data through an optimization algorithm; combining the output of the self-adaptive data quantification method and the input of the factor analysis method, and sharing learning information by the self-adaptive data quantification method and the factor analysis method through collaborative learning; a multi-task learning MTL algorithm is adopted, so that data quantification and factor analysis are mutually constrained in the training process; the quantitative standard and factor extraction are alternately optimized, so that the quantitative standard and factor extraction complement each other in the whole analysis process.

A wireless signal spectral analysis method using a multi-task learning-based spectral analysis learning model, the wireless signal spectral analysis method may be provided. The analysis method according to an embodiment of the present disclosure may include: receiving a target signal of a target band, obtaining a training dataset through pre-processing of the target signal, performing wireless signal spectral analysis learning using the training dataset, and analyzing the target signal using a trained spectral analysis learning model, wherein the performing of wireless signal spectral analysis learning comprises: configuring task specific layers for respectively performing individual learning for a plurality of tasks to be analyzed and a shared layer for performing shared learning; learning, in the shared layer, correlation data that meets a predefined criterion in the training dataset; and individually learning, in each of the plurality of task specific layers, using an individual dataset required for each task in the training dataset and a result of learning the correlation data.

The embodiment of the present application provides a river water quality detection method and system based on artificial intelligence, which significantly improves the accuracy and efficiency of water quality detection by obtaining on-site collected data of the river water quality detection system as regional water quality monitoring samples, and using a shared learning network to predict water quality status labels for these regional water quality monitoring samples. Specifically, by selecting the deep learning network with the best training effect as a shared learning network from multiple training entities participating in the linkage network parameter learning, the reliability and high precision of the water quality status label prediction data are ensured. Furthermore, based on these prediction data, the parameters of the target water quality assessment model are learned, and the optimized target application is determined, so that more accurate water quality status label prediction data can be generated. This method not only improves the intelligence level of water quality detection, but also effectively reduces the human errors and time costs that may exist in traditional water quality detection methods.

To reduce learning load on a learning model which controls a plurality of kinds of robots.SOLUTION: An operation learning device 1 for robots comprises: a plurality of first learning models 3a to 3c which accepts operation information at a certain time point and converts the operation information to feature quantities with respect to robots 2a to 2c; a shared learning model 5 which converts the feature quantities outputted from the first learning models 3a to 3c to predictive feature quantities at a next time point common to a plurality of kinds of robots 2a to 2c; a plurality of second learning models 4a to 4c which converts the predictive feature quantities at the next time point to predictive operation information with respect to the plurality of kinds of robots 2a to 2c; and a management unit 51 which uses teacher data relating to operations of the robots 2a to 2c to train the first learning models, the second learning models, or the shared learning model relating to the robots.SELECTED DRAWING: Figure 1

A robot motion learning device includes: a plurality of first learning models that receive motion information at a certain time and convert the motion information into features, for robots; a shared learning model that converts the features output by the first learning models into predicted features at a next time that are common to the plurality of types of robots; a plurality of second learning models that convert the predicted features at the next time into predicted motion information, for the plurality of types of robots; and a management unit that uses teaching data related to motion of the robots to train either the first learning model and the second learning model related to the robot or the shared learning model.