27 results about "Client data" patented technology

The application discloses a heterogeneous federated learning method based on space-time data distillation, which is used for solving the problem that related technologies cannot fully solve the time problem of knowledge reservation, resulting in knowledge forgetting when training a new global model, and the unique space-time characteristics of each client data are not fully considered, so that the efficiency and convergence of federated learning are poor. The server receives the local encoding data obtained by encoding based on the local data sent by each client for data-free global distillation, obtains a first global model, and distributes the first global model to each client; the client performs partial parameter local distillation and non-real-time distillation according to the local data and the first global model combined with dynamic temperature adjustment optimization, generates a local model, and sends the local model back to the server; the server updates the first global model according to each local model through a data-free global distillation strategy, obtains a second global model, and distributes the second global model to each client for next round training.

The application discloses an adaptive clustering federated learning modeling method for precision medicine, and relates to the technical field of precision medicine, and comprises the following steps: data collection and preprocessing and model construction and training; the application accurately determines the optimal number of the global model through adaptive clustering, and divides clustering groups according to the similarity of the client model parameters, realizes independent training of the groups, effectively improves the adaptability of the model to different client data characteristics, samples and label distribution differences, avoids the problem that a single model has poor performance in some clients, and significantly enhances the ability of the model to capture complex medical patterns; the model generalization is optimized through grouped training, so that the model can better cope with new data distribution, under the premise of ensuring the privacy and security of medical data, the accuracy and reliability of the model in precision medical scenes such as disease diagnosis and prognosis prediction are greatly improved.

This application proposes a data request processing method and an electronic device, relating to the field of application multi-version compatibility technology. The data request processing method includes: first, determining the application version information based on a received client data request; then, obtaining a target DSL configuration file for the application matching the version information; and finally, executing the data request using the target DSL configuration file to obtain the target data and return it to the client. This technical solution can improve the development speed of new versions and reduce the maintenance burden of multiple version applications within the server in scenarios requiring application multi-version compatibility.

Implementations described herein relate to receiving user input directed to an automated assistant, processing the user input to determine whether data from a server and / or third-party application is needed to perform certain fulfillment of an assistant command included in the user input, and generating a prompt that requests a user consent to transmitting of a request to the server and / or the third-party application to obtain the data needed to perform the certain fulfillment. In implementations where the user consents, the data can be obtained and utilized to perform the certain fulfillment. In implementations where the user does not consent, client data can be generated locally at a client device and utilized to perform alternate fulfillment of the assistant command. In various implementations, the request transmitted to the server and / or third-party application can be modified based on ambient noise captured when the user input is received.

The application discloses a kind of based on longitudinal federal learning's client data classification method, device and equipment, including obtaining to be detected client dataset, and to be detected client dataset is input to preset longitudinal federal learning classification model, preset longitudinal federal learning classification model includes feature coding module, feature purification module and server classification module;Data padding is carried out to to be detected client dataset using feature coding module, and output feature embedding dataset;Tensor decomposition is carried out to feature embedding dataset by feature purification module, and low-rank recovery tensor matrix is generated;Low-rank recovery tensor matrix is input to server classification module and is aggregated classification, and output target classification prediction result;The technical problem that the existing longitudinal federal learning data classification method can cause the situation that client data is missing in federal learning, thereby leading to the performance of longitudinal federal learning model is greatly reduced is solved.

This invention discloses a privacy-preserving robust federated learning aggregation method to resist backdoor attacks. Based on a dual-server architecture, clustering algorithm, and threshold filtering algorithm, it filters out malicious clients while ensuring the privacy of local updates by federated learning clients. First, this invention uses a dual-server architecture and, through arithmetic sharing and quadratic association pairs, enables the two servers to calculate the cosine distance between local updates of each client without knowing the plaintext of the client's local updates, thus achieving privacy protection for client data. Furthermore, this invention combines clustering operations and threshold filtering, allowing the servers to filter out updates from malicious clients based on both the direction and magnitude of the client's local updates, thereby improving the robustness of federated learning.

The present invention discloses a system and method for managing treatment plans and clinical documentation for substance use disorder (SUD) treatment. The system includes user devices associated with clients, clinicians, organization representatives, and supervisors. The system connects to a database storing client data, clinician data, assessment scores, and clinical documentation such as SOAP notes and progress notes. A computing device with artificial intelligence (AI) modules executes various program modules. A data collection module guides clinicians through predefined questions to gather information for generating SOAP and clinical progress notes. A secondary data collection module collects clarification from the counselor without requiring client input. A review module allows counselors to review, edit, approve, and submit documentation. A treatment plan generation module creates or updates plans based on collected data and prior scores. The system improves documentation efficiency, quality, and compliance, while supporting accurate and personalized treatment planning.

The application provides a kind of multimodal collaborative enterprise-level marketing content intelligent generation and distribution method and system, the method includes the following steps: S1, multimodal content data acquisition and structured processing are marked;S2, multimodal collaborative analysis and content performance feature extraction;S3, user portrait and enterprise portrait construction;S4, topic direction and content strategy generation;S5, multimodal marketing content generation;S6, distribution strategy matching and automatic distribution;S7, effect feedback and closed-loop optimization;By multimodal fine analysis to historical high-performance marketing content, and combining enterprise customer data and target user portrait, automatically generate targeted topic strategy and marketing content, while automatically distributing according to content features and platform constraints, and through effect data feedback to realize the continuous optimization of strategy and content, thereby solving the problem that marketing content generation in the prior art relies on artificial experience, low distribution efficiency and lack of closed-loop optimization.

This application belongs to the fields of artificial intelligence and fintech, and relates to an AI-based customer classification method, including: acquiring customer data and constructing a topological relationship network corresponding to the customer data; inputting the topological relationship network into a graph convolutional neural network for training to obtain a trained graph convolutional neural network; acquiring a target topological relationship network corresponding to target customer data, and processing the target topological relationship network through a customer segmentation model to generate segmentation results. This application also provides an AI-based customer classification device, computer equipment, and storage medium. Furthermore, this application relates to blockchain technology, and the customer segmentation model can be stored in the blockchain. The customer classification method of this application can be applied to user identification in the financial field, requires no labeled data for training, has low time complexity, and the use of the customer segmentation model can achieve fast and accurate generation of customer segmentation results.

Implementations described herein are directed to a framework for decentralized learning of large global machine learning (ML) model(s). In various implementations, remote processor(s) of a remote system can identify a global ML model, select client devices to participate in a given round of decentralized learning of the global ML model, and transmit, to each of the client devices, a processed version of the global ML model that is of a reduced transferrable size. Further, client device processor(s) of a client device can receive the processed version of the global ML model, obtain corresponding client data, perform partial model training, based on processing the corresponding client data, for the processed version of the global ML model to generate a corresponding update, and transmit the corresponding update back to the remote system. Moreover, the remote processor(s) can update, based on at least the corresponding update, the global ML model.

A network can comprise client devices, provider devices, and a database array. The network can be configured for RGB formatted data. A client device processes and converts client inputs into rows of client data columns in a synthesizing database stored in a memory device. The client data rows comprise transcoded RGB formatting of the client inputs. A provider device processes provider inputs. Either the database array or the provider device can process the provider inputs into rows of provider data columns of transcoded RGB formatting of the provider inputs. A collimation of the client and provider RGB formatted data can result in an indicia of complementary client and provider inputs.

The application discloses an enterprise client data service management system, comprising: a data acquisition layer, used for collecting data change events in real time from multiple source heterogeneous data sources; an intelligent computing layer connected to the data acquisition layer, used for processing and calculating data, comprising a Merkle tree-based incremental synchronization module and an adaptive multi-engine query routing decision module; a service encapsulation layer connected to the intelligent computing layer, comprising a Lisp dialect-based unified index definition and translator, used for encapsulating business index definition as a standard data service API; and a unified access layer, providing a unified API gateway for external applications, used for receiving query requests and returning processing results. Through real-time acquisition and streaming calculation, the application shortens the data service response time from hours to seconds, meeting the stringent requirements of business on data timeliness. Implementation cases show that the time consumption of ten million data export is shortened from more than 30 minutes to less than 3 minutes, and the performance is improved by more than 10 times.

This application provides a method and apparatus for training a diversity data aggregation model based on industrial network partitioning, relating to the field of computer technology. It enables model training by adjusting client partitions according to data diversity, thereby improving data security and robustness during model training. The method includes: partitioning multiple clients into industrial network regions based on client characteristics and robustness parameters of each client, obtaining client regions, each client region including at least one client; acquiring client data from each client region; training the model based on client data to obtain regional model parameters corresponding to each client region; aggregating the regional model parameters from each client region to obtain global model parameters; and updating the model parameters based on the global model parameters.

The application provides an AI real-time early warning method based on a vehicle networking JTT808 protocol, comprising the following steps: S1: log collection and E2E-ID identification extraction; S2: link aggregation and bidirectional association; S3: node index storage and link index aggregation analysis; S4: AI early warning analysis; S5: client data visualization interaction and display; the application realizes automatic bidirectional association of uplink and downlink by generating a globally unique E2E-ID and transmitting and inheriting, combines with an associated ID to form a unified full-link track, does not need manual association, and improves fault troubleshooting efficiency; meanwhile, probes are deployed at four types of nodes to collect fine-grained data, a node list and an edge list are generated, end-to-end node observation is realized through multi-view visualization, and an abnormal node is accurately located.

This invention provides a federated learning method using synonymous data, comprising: a coordinating device sending a general model to each client device; each client device executing a training procedure, including: an encoder encoding private data into a summary; training a client model based on the private data, the summary, and the general model; and sending the summary and client parameters of the client model to the coordinating device; the coordinating device identifying absent client devices among the client devices; generating synonymous data using a synonymous data generator based on the summary corresponding to the absent client device; training an alternative model based on the synonymous data and the summary corresponding to the absent client device; and performing an aggregation operation based on the alternative model parameters and the client parameters of each client device other than the absent client device to generate update parameters to update the general model; this invention addresses the problem of client departure by synthesizing representative client data in a coordinator.

This document describes a transmission network, network elements, and usage methods, including a transmitter comprising a client-side input, a transmitter circuitry, and an antenna. The client-side input is configured to receive a baseband signal encoded with client data. The transmitter circuitry is configured to receive the baseband signal from the client-side input and generate an antenna feed signal based on the baseband signal. The antenna is configured to receive the antenna feed signal from the transmitter circuitry, generate a radiated signal based on the antenna feed signal, and couple the radiated signal into a hollow waveguide. Each radiated signal is a radiated electromagnetic wave configured for coherent detection and having a frequency in the range of 300 GHz to 10 THz.

The application discloses a UAV nest cooperative operation method and system based on personalized federated learning, and belongs to the technical field of edge computing and cloud computing cooperation. The method comprises the following steps: collecting original operation and maintenance data of a plurality of UAV nests, wherein the original operation and maintenance data comprises at least one sensor reading and a corresponding scene label; performing scene clustering processing on the original operation and maintenance data to generate a scene clustering cluster; calculating the feature weight of each operation and maintenance feature; weighting and fusing the client data in the scene clustering cluster by using the feature weight to generate weighted training data, and training a scene-specific basic parameter on the edge server side; uploading the scene-specific basic parameter to a cloud server for federated aggregation to generate global model parameters; distributing the global model parameters to the client, and performing personalized updating on the global model parameters based on local data on the client to generate personalized parameters; and performing a UAV nest operation and maintenance prediction task based on the personalized parameters.

This invention proposes a personalized Bayesian federated learning method and system based on conjugate multi-class Gaussian process classification and deep kernel learning, belonging to the field of federated learning technology. The invention discloses a personalized Bayesian federated learning framework (FedCMGP) based on conjugate multi-class Gaussian process classification. This framework achieves analytical inference of the posterior of multi-class Gaussian processes by combining One-vs-Each approximation and Pólya-Gamma augmentation techniques on the client side; on the server side, it collaboratively learns shared priors through deep kernel learning and uses a federated averaging algorithm for global aggregation. This invention aims to address the technical shortcomings of existing federated learning methods in handling non-independent and identically distributed client data, quantifying prediction uncertainty, reducing communication overhead, and maintaining robustness in noisy or adversarial environments, providing a more accurate, reliable, and communication-efficient solution for distributed machine learning.

This invention discloses a verifiable heterogeneous federated learning system based on zero-knowledge proofs, primarily addressing the problems of insufficient transparency, lack of verifiability, and vulnerability to data poisoning attacks in existing heterogeneous federated learning systems. The system consists of multiple client devices with different computing capabilities, data distributions, and model architectures, and a blockchain verification platform. The client devices are responsible for local model training, model conversion to the standard ONNX format, generating proofs using zero-knowledge proof tools, and submitting model updates and proofs to the blockchain verification platform. The blockchain verification platform is responsible for verifying zero-knowledge proofs, recording model updates, constructing a global model, and ensuring its verifiability and transparency. This invention effectively protects client data privacy, enhances the credibility of the training process, ensures the verifiability of the global model, provides an enhanced path for decentralized heterogeneous federated learning, and expands new directions for the development of trusted federated learning.

A computing apparatus includes a processor, a communication interface coupled via a bus configured to communicate over a communication network with provider computing devices, external data server devices, and clinical server devices, and a memory storing a decision engine, workflow engine, outcome analysis tool, parameter recommendation tool, curriculum rules database, and hours rules database. The processor receives electronic requests, retrieves client specific data from external data server devices, and analyzes the data using automated evaluation assessment tools. The decision engine correlates client specific data with prior stored client outcome data to identify matching data patterns, improving accuracy. The workflow engine automatically processes and correlates evaluation data with stored curriculum and hour rules from clinical server devices to determine parameters without manual intervention, reducing processing time. Plan data is generated by correlating scores with client data stored in memory and transmitted over the communication network, standardizing data processing across multiple provider computing devices.

The disclosure relates to a system for automating a request for proposal (RFP) process. In one embodiment, the system includes a SmartRFP software tool for initiating and managing response process through a workflow, processing an RFP document to generate an RFP response document. The SmartRFP software tool comprises a user interface, a generative artificial intelligence (GenAI) engine, a content management system and a dashboard reporting function. The user interface receives service requests, client profile data, and manages processing of the RFP document. The GenAI engine extracts content from the RFP document, identifies questions in the RFP document, and generates answers based on the identified questions using an authoritative repository. The GenAI generates the answers by decomposing the identified questions into multiple sub-questions, rephrasing each sub-question, and classifying each rephrased sub-question into one or more topics. The content management system manages and builds the content and stores and retrieves the content from the authoritative repository in collaboration with the GenAI engine. The reporting dashboard renders performance and business insights data.

The application discloses a multi-queue data packet asynchronous distribution method and system based on atomic operation, and the method comprises the following steps: a server receives an original data packet and calculates a total length of the data packet; a current head pointer in a ring buffer is read through atomic operation, and a next writing position is calculated according to the current head pointer and the total length of the data packet, and the data is written into the ring buffer; the head pointer is updated through atomic operation by the server, and the current writing position is recorded; the ring buffer stores the original data packet written by the server and the head pointer of the server through a data storage area of shared memory; the data storage area is a pre-allocated fixed-size ring memory area; each client reads the current head pointer, the data packet head and the original data packet from the ring buffer and parses and processes the same, and each client has and maintains an independent tail pointer, and records the self-consumption progress. The application realizes efficient asynchronous distribution of multi-client data packets through the shared memory ring buffer and atomic operation.

Transmission networks, network elements, and methods of use are described herein, including a transmitter comprising a client-side input, a transmitter circuit, and antennas. The client-side input is configured to receive baseband signals having client data encoded therein. The transmitter circuit is configured to receive baseband signals from the client-side input and generate antenna feed signals based on the baseband signals. The antennas are configured to receive antenna feed signals from the transmitter circuit, generate radiated signals based on the antenna feed signals, and combine the radiated signals within a hollow waveguide. Each of the radiated signals is a radiated electromagnetic wave configured for coherent detection and has a frequency within the range of 300 gigahertz (GHz) to 10 terahertz (THz).

This invention provides a compatible aggregation method and system for multi-source heterogeneous microservice API documents. The system includes a multi-channel service discovery module, a specification detection and version identification module, a heterogeneous document conversion engine, an aggregation storage and caching module, and a version-aware front-end display module. The method includes the following sub-steps: S1: Parallel execution of service collection operations to obtain service instance information; S2: Interface description document version detection and specification identification; S3: Unified conversion of heterogeneous interface documents; S4: Aggregation storage and caching; S5: Client data display and interface retrieval. This method accesses unregistered services through static configuration and manual registration. It is suitable for legacy systems not connected to the registry center, external third-party services, and temporary services in the development environment. It achieves low-intrusion access without modifying service code, deploying additional proxies, or adjusting the network architecture.

This invention discloses a federated training scheduling method based on hierarchical dynamic selection and diversity constraints, belonging to the field of distributed machine learning. This invention aims to address the problems of low training efficiency and unstable convergence in existing federated learning methods under environments with both statistical and system heterogeneity. This invention obtains a set of available clients, maintains latency estimates for each client and hierarchically divides them, determines the sampling quota for each layer based on the average latency, historical participation, and fairness constraints, constructs a client data difference graph within each layer, and uses a greedy decentralized selection strategy with diversity constraints to determine the set of participating clients. Local updates for selected clients are aggregated and the client states are updated. This invention is applicable to distributed training scenarios such as mobile terminal collaborative learning, vehicle networking, and edge intelligent analysis.