Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

88 results about "Model aggregation" patented technology

Equipment evaluation and federated learning importance aggregation method, system and equipment based on edge intelligence and readable storage medium

The invention provides an equipment evaluation and federated learning importance aggregation method based on edge intelligence, which comprises the following steps of cloud server initialization: generating an initial model by a cloud server, equipment evaluation and selection: receiving resource information of terminal equipment by an edge server, generating a resource feature vector, and inputting the resource feature vector to the evaluation model, local training: after the edge server selects the intelligent equipment, sending the transferred initial model to the intelligent equipment, andenabling the intelligent equipment to carry out local training on the initial model in federated learning to obtain a local model, local model screening: sending the local model to an edge server, and judging whether the local model is an abnormal model or not by comparing the loss values of the local model and a previous round of global model, and global aggregation: performing global aggregation by using a classical federated average algorithm. According to the method provided by the invention, on one hand, the training bottleneck problem with resource constraint equipment is solved, and onthe other hand, the model aggregation effect is improved so as to reduce redundant training and communication consumption.
Owner:HUAQIAO UNIVERSITY

Random convolutional neural network-based high-resolution image scene classification method

The invention discloses a random convolutional neural network-based high-resolution image scene classification method. The method comprises the steps of performing data mean removal, and obtaining a to-be-classified image set and a training image set; randomly initializing a parameter library of model sharing; calculating negative gradient directions of the to-be-classified image set and the training image set; training a basic convolutional neural network model, and training a weight of the basic convolutional neural network model; predicting an updating function, and obtaining an addition model; and when an iteration reaches a maximum training frequency, identifying the to-be-classified image set by utilizing the addition model. According to the method, features are hierarchically learned by using a deep convolutional network, and model aggregation learning is carried out by utilizing a gradient upgrading method, so that the problem that a single model easily falls into a local optimal solution is solved and the network generalization capability is improved; and in a model training process, a random parameter sharing mechanism is added, so that the model training efficiency is improved, the features can be hierarchically learned with reasonable time cost, and the learned features have better robustness in scene identification.
Owner:WUHAN UNIV

Virtual power plant combined heat and power scheduling robust optimization model

The invention provides a virtual power plant combined heat and power scheduling robust optimization model. A model aggregation unit comprises a distributed generating set, a wind turbine generator set, a photovoltaic set, a combined heat and power (CHP) set, a boiler, electric energy storage, heat energy storage, an electric load and a heat load. Participation of the CHP set in the SRM (Spinning Reserve Market) situation is considered. Aiming at the facing uncertain problem of a virtual power plant (VPP) and resulting risks, robust optimization (RO) is utilized to process uncertainty of the EM electricity price, the SRM electricity price, the wind power capacity, the photovoltaic capacity, the electric load and the heat load, and risk quantification indexes are established, and thus robustness and economical efficiency of a RO model are balanced. The model provided by the invention well solves the existing combined heat and power scheduling optimization model establishment problem of the VPP when participating EM and SRM at the same time, and improves flexibility of decision making, and thus the profit of the VPP is increased. Meanwhile, the introduction of the RO model effectively reduces system risks, and thus the effective reference is provided for a decision maker to select a proper robust factor.
Owner:HOHAI UNIV

Virtual power plant day-ahead scheduling optimization model

InactiveCN106169102ASolve the problem of dealing with the uncertainty of wind power output at the same timeIncrease profitForecastingSystems intergating technologiesDecision schemeOptimal decision
The invention provides a virtual power plant day-ahead scheduling optimization model. A model aggregation unit comprises a gas turbine, a wind turbine generator set, a photovoltaic set, a water drawing energy storage power station and loads. For the characteristics that the electricity price probability distribution description is relatively accurate and the prediction is relatively high, random programming is adopted to process the uncertainty of the electricity price; and for the characteristics that the wind power and photovoltaic output probability distribution is difficult to precise describe and the prediction precision is relative low, an information gap decision theory (IGDT) is adopted to process the uncertainty of wind power and photovoltaic output, different weights are provided to wind power and photovoltaic output deviation coefficients, and the IGDT is enabled to simultaneously process the uncertainty of wind power and photovoltaic output. In addition, for the blindness of uncertainty decisions and the different risk degrees of different strategies, the risk cost is introduced, and the risks corresponding to different decision schemes are quantified. According to the invention, a larger decision making space is provided for a decision maker, and the VPP is enabled to make the optimal decision under more conditions, so that the benefit of the virtual power plant (VPP) is increased.
Owner:HOHAI UNIV

Method and system for establishing user behavior periodic model based on transverse federated learning

The invention provides a method and system for establishing a user behavior period model based on transverse federated learning. The method comprises steps of collecting the historical behavior data of a user, and carrying out the preprocessing of the historical behavior data of the user; dividing the preprocessed data according to user IDs, and constructing time series data; performing vectorization processing on the time series data; transverse federation learning being carried out according to the user feature vectors, each participant training a local model locally by using a long-term andshort-term memory artificial neural network, and trained model parameters being uploaded to a model aggregation server; updating and aggregating the model parameters to obtain aggregated model parameters; detecting whether the aggregation model is converged or not, and if not, returning aggregation model parameters to the participant to continue iterative training until the detected aggregation model is in a convergence state; and establishing a user behavior period prediction model according to the aggregation model parameters of the aggregation model in the convergence state, and downloading the user behavior period prediction model to the participant.
Owner:BANK OF CHINA

User customer group classification method and device

The invention provides a user customer group classification method and device, and the method comprises the steps: obtaining user feature data in participation nodes of a federated learning distributed network, wherein the federated learning distributed network comprises participation nodes and model aggregation nodes; training a logistic regression model in the participation nodes according to the user characteristic data, and determining participation node gradient ciphertext information; uploading the participation node gradient ciphertext information to a model aggregation node of the federated learning distributed network for aggregation, and determining aggregation gradient ciphertext information; according to the aggregation gradient ciphertext information, carrying out transverse federated learning in model aggregation nodes, and determining joint gradient information; distributing the joint gradient information to each participation node, and inputting the joint gradient information to a federated learning logistic regression customer group classification model for training; and classifying the user customer groups according to the trained federated learning logistic regression customer group classification model. According to the invention, the accuracy of user customer group classification can be improved.
Owner:BANK OF CHINA

Block chain cross-chain-based federated learning method and device

The invention discloses a block chain cross-chain-based federated learning method and device, and the method comprises the steps: carrying out the training of a federated learning model through local data in a single block chain network, and transmitting model parameters to a model aggregation smart contract through a private transaction, thereby achieving the aggregation of different node model parameters in a single chain, synchronizing model parameters at each node; the model aggregation smart contract sending the latest model parameters to a cross-chain network through cross-chain private transaction to realize model synchronization between different chains. In the whole process, data of a single network node and data of a cross-chain network node are not exchanged, it is ensured that model parameters of each node are not leaked through a private transaction mode, training of different node data of different block chain networks on a federated learning model is achieved while data privacy security is ensured, a trained data set is expanded, and the accuracy of the model is improved. An integral mechanism is adopted to improve the enthusiasm of each member to contribute to a data training model, and the model training effect is further improved.
Owner:CHINA ZHESHANG BANK +1

Decentralized federated learning framework based on heterogeneous computing power perception and modeling method

The invention discloses a decentralized federated learning framework based on heterogeneous computing power perception. The decentralized federated learning framework comprises a cloud coordinator and a plurality of equipment ends. The cloud coordinator is used for management, training and parameter updating scheme generation during operation and regular model backup. The equipment end is used for transmitting equipment information to the cloud coordinator, operating a model locally and updating equipment end parameters. The cloud coordinator acquires the least common multiple of the one-time training time of the equipment end as a super cycle, the equipment end calculates different step lengths in the super cycle, and the model is aggregated in the integral multiple of the super cycle. Different local steps are operated according to different computing capabilities of equipment, and in the model aggregation process, negative effects of slow nodes are reduced; and a distributed point-to-point communication mode is adopted, and the communication pressure of the central server in the distributed training process can be eliminated under the condition that the overall communication traffic is not increased.
Owner:SUZHOU INST FOR ADVANCED STUDY USTC

Heterogeneous model aggregation method and system based on federated learning

The invention relates to the field of federated learning, in particular to a heterogeneous model aggregation method and system based on federated learning, and the method comprises the steps of initializing a neural network model; the method also includes that each client contributes a part of local data and uploads the local data to the server to form a shared data set, and a CGAN model is trained; the client uses a local data set and a data set generated by the CGAN model to train a local model, predicts each data in the shared data set and uploads a prediction score to the server; the server calculates the prediction score deviation degree of each client, takes the reciprocal of a calculation result as a weight, calculates a global prediction score, and uses the global prediction score to perform knowledge distillation on the server model; the client downloads the prediction scores of other client models from the server for cooperative training; and model convergence is performed after multiple iterations. According to the invention, the problem of data heterogeneity of the client side can be solved, the client side model uploads and downloads the prediction score of the shared data set, and the communication traffic between the client side and the server side is reduced.
Owner:GUANGZHOU UNIVERSITY

Federal learning method with high communication efficiency in wireless communication scene

The invention discloses a federated learning method with high communication efficiency in a wireless communication scene. The method comprises the following steps: S1, constructing a federated learning framework from three aspects of a federated learning system, a training algorithm and a communication model in the wireless communication scene; s2, aiming at the constructed federal learning framework, carrying out convergence analysis on the training process of the federal learning framework; and S3, constructing an optimization problem about the federated learning framework according to a convergence analysis result, and solving the problem through a joint optimization method for equipment selection and beam forming. According to the method, on the basis of air calculation and a second-order training algorithm, on one hand, low-delay model aggregation is achieved through the waveform superposition characteristic of a channel, on the other hand, the number of iteration rounds needed by training is reduced through the rapid convergence characteristic of the second-order algorithm, and the communication bottleneck problem existing in most wireless federated learning methods at present is solved. Meanwhile, the training accuracy is further improved through the provided joint optimization method for the federated learning framework.
Owner:EAST CHINA NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products