137 results about "Supplementary data" patented technology

The invention relates to the technical field of data processing and financial science and technology, and particularly discloses a vehicle insurance claim data processing method and device, a storage medium and computer equipment. The method comprises the following steps: acquiring vehicle insurance claim settlement application data corresponding to a vehicle insurance claim settlement behavior, vehicle feature supplementary data of a claimed vehicle and user feature supplementary data of a claim settlement user; according to the vehicle insurance claim settlement application data, the vehicle feature supplementary data and the user feature supplementary data, generating a vehicle insurance claim settlement feature vector, and based on the vehicle insurance claim settlement feature vector, determining a first fraud probability through a preset fraud behavior identification model; according to the vehicle feature supplementary data and the user feature supplementary data, constructing a multi-subject claim relation graph, and based on the multi-subject claim relation graph, determining a second fraud probability corresponding to the claim user; and determining a target fraud probability of the car insurance claim settlement behavior according to the first fraud probability and the second fraud probability.

There is provided a data analytics function comprising a processor and a receiver. The processor is arranged to generate analytics data for an analytics service using at least one data source. The receiver is arranged to receive an event related to the analytics service. The processor is further arranged to determine a rating of the at least one data source, the rating based on supplementary data.

This invention discloses an EMC visualization guidance method based on EMC data assets, belonging to the field of EMC design assistance technology. It includes: identifying the visualization object and obtaining its identification information; retrieving matching quantitative data using the identification as an index, generating a supplementary data request form when data is missing; mapping the quantitative data into expert-style graphical elements to generate a visualization guidance interface, progressively displaying basic physical concepts, coupling paths, three-layer protection, test types, process assets, and decision reporting views; receiving user interaction commands, recording the operation sequence and corresponding visualization status during historical operation sequence playback, forming an operation sequence and status snapshot data asset and assigning multi-dimensional searchable tags; and generating visualization output based on this data asset. This invention transforms abstract EMC data into intuitive graphics, records the key evolution of the design and diagnostic process, supports timeline playback, analog guidance, and cross-domain reuse, realizing the digital accumulation and efficient reuse of design experience.

This invention relates to a method and apparatus for detecting and correcting federated learning data. The method includes: acquiring intermediate data uploaded by each participating party, the intermediate data including location information of data to be supplemented, suspicious data, and indicators related to the suspicious data; performing a global analysis based on the indicators related to the suspicious data to filter the suspicious data and determine the data to be corrected; determining a problem data matrix based on the location information of the data to be supplemented and the data to be corrected; acquiring the associated data corresponding to each problem data in the problem data matrix from each participating party, analyzing the associated data, and determining a correction matrix for each participating party, so as to provide data processing services based on the correction matrix of the participating parties and the local data of the participating parties. The technical solution provided by this invention can effectively solve the technical problems of data accuracy, consistency, and security by combining data quality, data validity detection, and data protection mechanisms.

The invention relates to a resource scheduling method and system based on a large language model, a readable medium and equipment, and relates to the technical field of cloud computing and artificial intelligence. According to the scheme, a cluster is connected, and a scheduling analysis task is constructed in response to a first scheduling request initiated by a scheduler at the beginning of a target container scheduling period; based on the prompt template, constructing inference prompt words and inputting the inference prompt words into a large language model; obtaining session supplementary data updating reasoning prompt words based on a calling instruction submitted by a large language model; when a resource scheduling suggestion submitted by the large language model is received, the scheduling analysis task is ended, and the resource scheduling suggestion is provided; the session supplemental data may include real-time resource status of the cluster, corresponding approximate case information in historical knowledge. According to the scheme, in the resource scheduling analysis process, thinking feedback of a large language model is used for calling real-time resource dynamic and historical approximate cases as supplementation, and therefore resource scheduling requirements are dynamically analyzed; and the separation of knowledge and reasoning is realized, and the maintenance and expansion cost is reduced.

The invention discloses a target user-oriented comprehensive energy technology matching degree intelligent recommendation system, and particularly relates to the technical field of comprehensive energy technology matching. The system firstly constructs a double-clock data base with a unified time base and a fixed step length, governs late and additionally recorded data and generates an evidence pointer; consistency and traceability of measurement calibers of electricity, heat, cold, gas, steam and the like are ensured. And then performing same-caliber linkage with exogenous factors such as meteorology, carbon factors and policies on the quality-weighted user portraits to form a scene library which can be used for future deduction. Generating candidate schemes based on a technical library, executing coarse-to-fine system-level feasibility check of multi-energy coupling and engineering implementation constraints, and performing quasi-dynamic verification in a prediction window to obtain a feasible scheme set and binding a key state quantity interval of the feasible scheme set with evidence; and finally, calculating a comprehensive matching degree and a confidence interval by using a multi-objective function of scene probability and risk penalty, and outputting a scheme and a strategy.

An emergency response system provides a map-based interface for a telecommunicator to view information about an incident and coordinate a response to the incident. The emergency response system gathers supplemental data regarding locations and other information that may be relevant in assisting with the incident. The interface may automatically select relevant supplemental information based on the incident and provide this incident-specific information for display on the interface. The user may then select a response on the interface such as a unit to dispatch that the system automatically implements, providing unified information and control for incidents, supplemental information, and incident response.

The present disclosure provides a training method for a machine learning model and a refinement method for training samples. The training method includes: obtaining feature vectors of the training samples, clustering the feature vectors to obtain representative training samples, and then training the machine learning model based on the representative training samples. The refinement method includes: querying an external database based on an original sample to obtain supplementary data, using a machine learning model to evaluate the original sample to generate review data, and then using another machine learning model to refine the original sample based on the supplementary data and the review data.

The invention relates to a real-time three-dimensional modeling system fused with unmanned aerial vehicle target recognition, and relates to the technical field of unmanned aerial vehicle application, and the system comprises a data collection module which is used for obtaining multi-source sensor data; the real-time modeling module is used for constructing a real-time three-dimensional model of the environment The target identification module is used for identifying a specific target in the scene and associating the specific target with the three-dimensional model. The information quality evaluation module is used for quantifying the observation information quality of the recognized target in real time according to factors such as the shielding degree, the resolution ratio and the observation angle; and when the information quality evaluation result is lower than a preset threshold value, the flight path planning module is triggered to autonomously plan a dynamic flight path going to the optimal viewpoint, and the flight control module executes the flight path and drives the unmanned aerial vehicle to perform supplementary data acquisition. According to the method, the autonomous closed loop of the perception evaluation decision action is established, so that active and intelligent fine modeling of the key target is realized, and the integrity of model information and the operation efficiency are improved.

The application relates to a non-IID federated learning model backdoor robustness evaluation method based on a generative adversarial network, which comprises the following steps: selecting a client in federated learning as a test node, taking a downloaded server-side global model as a discriminator, and designing a generator locally to form a generative adversarial network model; a backdoor attack target category is specified, a class representative sample is reconstructed by using the generator in each round of global training, and the target category is marked as pre-poisoning data participating in training; a supplementary data set is generated locally; a source category of the backdoor attack is specified, a backdoor trigger is optimized, the supplementary data set is used for class-specific backdoor training, the discriminator is updated and uploaded to the server, and the global model is updated; the non-IID degree of data and the number of specified source categories are adjusted, and the robustness of the federated learning global model to the backdoor attack is observed. Compared with the prior art, the application can verify the effect of the backdoor attack on the federated learning model under different degrees of data heterogeneity.

This invention provides a method and system for remote meter reading and water usage information management for smart water meters. The method mainly includes: receiving raw data frames containing information such as cumulative flow values ​​and collection timestamps, and performing multi-dimensional quality checks to generate collection status identifiers; dividing the water usage interval into valid collection intervals and pending confirmation intervals based on the identifiers; calculating the measured water usage in the valid intervals and extracting abnormal features to generate initial abnormal events; generating a completion sequence for the pending confirmation intervals based on boundary flow values, historical water usage characteristics, and regional balance constraints; determining a supplementary reading control strategy and issuing supplementary reading instructions based on the completion sequence and other information; performing secondary verification of the initial abnormal events based on the supplementary data, and outputting the final abnormal result and meter reading data. This method improves the accuracy of water usage anomaly detection and the reliability of meter reading data through data quality grading, interval hierarchical management, and closed-loop verification of abnormal events.

Systems and methods are disclosed for attributing advertisement exposure in a cinema environment. A ticket event is accessed, indicating a user's purchase of a movie ticket at a specific location, screening room, and showtime. Advertisement schedule data is accessed to identify ads shown before the movie. The system determines that the user was likely present during the advertisement window using ticket information and supplemental data such as geolocation or concession purchase records. Post-exposure events—such as transactions, store visits, or digital engagement—are analyzed to detect behavioral changes following ad exposure. A correlation engine links the exposure to user behavior, enabling ad attribution, effectiveness scoring, and campaign optimization. The system supports privacy-preserving data handling and applies statistical or machine learning models to improve attribution accuracy. This enables advertisers to measure the impact of cinema-based advertisements and make data-driven decisions across physical and digital channels.

The application discloses a power grid inspection method and device, and relates to the technical field of electric power. The method comprises the following steps: detecting whether a user instruction is received; if the user instruction is received, determining a communication protocol of each inspection point in the power grid based on a distributed service architecture, and collecting power grid data through the communication protocol; processing missing values and abnormal values in the power grid data, and integrating the processed power grid data through supplementary data to obtain integrated power grid data; predicting the integrated power grid data based on a prediction model to obtain an analysis result; wherein the analysis result at least comprises a comprehensive operation state of the power grid; the prediction model is a data analysis model constructed based on a machine learning algorithm; determining an inspection strategy and executing the inspection strategy according to the analysis result; wherein the inspection strategy comprises an optimal inspection path; and the optimal inspection path is obtained based on an inspection time, energy consumption and equipment health state prediction through a path planning model. The application is used for realizing the inspection function of the power grid.

This invention provides a method and system for automatic querying and processing of list data based on intelligent algorithms, relating to the field of data query technology. The method involves acquiring the list data to be queried and its heterogeneous modal supplementary data, modeling them into a unified heterogeneous semantic graph, performing multi-hop reasoning on the graph to obtain semantic paths, converting the paths into query operation sequences using a policy decoder, and performing legality verification by combining structural and semantic constraints. Finally, a query strategy sequence that satisfies both structural legality and semantic consistency is generated, improving the accuracy and intelligence level of list data querying.

The invention discloses a training data balance processing method and device, equipment, a medium and a product. The invention relates to the technical field of data processing. The method comprises the following steps: when a user in a user set is associated with an object in an object set, generating associated data according to the associated user and object, and adding the associated data into an associated data set; acquiring other users of the same category as the target user corresponding to the to-be-supplemented data of the associated data set; obtaining a target object associated with the target user in the associated data set; according to each target object, determining candidate objects of the target user in other objects associated with the other users in the associated data set; and generating associated data according to the target user and each candidate object, and adding the associated data into the associated data set. The embodiment of the invention can balance the training data of the model.

Embodiments of the present application provide an order data processing method and device, a storage medium and a computer program product. The method applied to a target platform comprises: receiving an order data processing request, the order data comprising order payment completed outside the target platform; obtaining supplementary data in the order completion process according to the order data; and determining that the order payment is completed outside the target platform according to an identification result of the supplementary data. Therefore, the embodiments of the present application can know whether the user completes the order payment outside the platform without manual customer service processing, and improve the processing efficiency of feedback.

A mobile agricultural machine receives a topographic map indicative of topographic characteristics of a worksite, wherein the topographic characteristics are based on data collected at or prior to a first time and receiving supplemental data indicative of characteristics relative to the worksite, the supplemental data collected after the first time. A topographic confidence output is generated which is indicative of a confidence level in the topographic characteristics of the worksite as indicated by the topographic map, based on the topographic map and the supplemental data. In some examples, an action signal is generated to control an action based on the topographic confidence output.

The invention discloses a service full-link real-time perspective and driving method, device and equipment and a storage medium, and the method comprises the steps: building a hybrid data integration system, and integrating data dispersed in each heterogeneous service system to a lightweight order center; obtaining a data change event of the core business system through database log mining and issuing the data change event to a message queue; updating the lightweight order center based on the incremental data; calling a source system interface to acquire and process missing data in real time during query through an on-demand back-to-source mechanism; assembling the basic data and the supplementary data into a complete business object, and generating a full-link state view through state mapping; and when the key service state node is identified to be changed, broadcasting an event to a subscription system to trigger downstream service process automation. According to the invention, real-time integration and second-level response of service data are realized, data blind areas are eliminated while stability is guaranteed, full-link state perspective and service process automation are realized, and service operation efficiency and reliability are improved.

The application provides a trackless navigation texture information redundancy analysis method, comprising: processing original image data of a current wear area by using an edge detection algorithm to obtain wear boundary contour coordinates and gray distribution data; extracting boundary coordinate point sets from a reliable anchor point area and determining an optimized update frequency value according to the position and shape matching degree of the boundary coordinate point sets and historical path map data; incrementally updating the path map by using the optimized update frequency value, integrating the gray distribution data of the reliable anchor point area and the texture detail loss degree value, and generating an enhanced positioning reference data set; updating the enhanced positioning reference data set according to supplementary data and performing periodic matching degree evaluation according to the optimized update frequency value to maintain long-term positioning consistency.

System and method secures critical data on distributed ledger (DL) or blockchain (Bchn) or linked memory cell (LCM) data construct. Each cell in DL / Bchn / LCM has cellular immutability and cellular connectivity characteristics. System randomly stores critical data in the DL / Bchn / LCM based with initial state function and stores nonsensical data in other cells based upon nonsensical initial state function which is inferior to the critical data function. The DL / Bchn / LCM has a stem, with stem cells, forking cells and branches with branch cells created by forking initial state function (sometimes triggered by cyber attacks, sometimes by new growth). Trees of DL / Bchn / LCM are formed as are forests of DL / Bchn / LCM trees. automatic information search operations (AISO) through data collections also grow stems and branches on the DL / Bchn / LCM trees. The AISO generates supplemental (Supple) data based on supple initial state function. Supple data is further dispersed and randomly stored in the DL / Bchn / LCM trees and / or forest.

The application discloses a kind of data analysis methods of intelligent police law enforcement glasses, belong to data analysis technical field, comprising: extracting video stream from multiple source data sets for identification, determine whether the identification result satisfies preset threshold, if satisfy then the target feature is as video feature set, if not satisfy then trigger supplementary data acquisition and combine original target feature to obtain video feature set;Key elements in video feature set are spatiotemporally aligned with voice instructions, to obtain comprehensive event description vector;Sequence analysis is carried out on comprehensive event description vector to obtain event trend vector, to obtain event analysis result;According to priority, real-time voice feedback content is obtained by converting event analysis result;The voice broadcast signal of real-time voice feedback content is generated.The data analysis method of the intelligent police law enforcement glasses solves the problem that information lags due to insufficient data processing speed, which makes it difficult to ensure the efficiency and safety of the law enforcement process.

A method of identifying a care gap for a patient comprising: inputting identification data relating to a patient into a clinical decision engine; retrieving, from an FHIR repository, a supplementary d

Disclosed embodiments relate to systems and methods for detecting and mitigating unwanted instructions in artificial intelligence models authentication at a client application. Example techniques include receiving, an original request from a user; identifying historical data associated with a previous interaction with the user; receiving, from a trained model, an action identifier associated with raw data stored in a data source; receiving the raw data from the data source as a result of performing an action based on the action identifier; supplementing the raw data with a relevant portion of the historical data to generate supplemented data; determining whether the supplemented data includes an instruction; and based on a determination that the supplemented data includes the instruction, performing a control action.

The invention discloses a multi-source meteorological data fusion processing method for new energy output prediction, and the method comprises the steps: S1, determining a new energy type and a core meteorological impact factor, and determining the data time granularity according to a prediction time scale; s2, collecting localized observation, reanalysis and numerical forecasting data, removing abnormal values and unifying the abnormal values into a structured format; s3, unifying data resolution through bilinear interpolation, and realizing meteorological data and station space matching in combination with station boundary cutting and space verification; s4, unifying all data as Beijing time, aligning timestamps, filling missing values according to priorities, and controlling a single factor missing rate to be less than or equal to 5%; s5, quality control is carried out from the dimensions of data integrity rate, spatial correlation and the like, and if the data does not reach the standard, an interpolation method is adjusted or a data source is supplemented; s6, outputting CSV or netCDF format fusion data according to requirements, and performing classified storage according to stations and time so as to facilitate subsequent association; the method has the advantages of being good in input data basis, good in output data space-time consistency and convenient to use.

Method for determining damage to a vehicle using an optical detection unit for three-dimensional detection of a vehicle surface and an evaluation unit connected to the optical detection unit, with which data of the three-dimensionally detected surface can be processed and spatial coordinate data of the three-dimensionally detected surface can be determined, characterized in that the evaluation unit has a database in which vehicle model data, comprising vehicle type data, standard spatial coordinate data of the vehicle type, a process data record comprising damage data, repair data and optionally process supplementary data, wherein the damage data includes vehicle type data, area coordinate data of the spatial coordinate data of a damaged area of ​​the surface of a vehicle, deformation depth data, wherein the deformation depth data exhibits the largest negative difference between the depth coordinate data of the spatial coordinate data in the damaged area and the standard spatial coordinate data stored in the database, wherein the Repair data, spare parts data and working time data, and optional additional process data, such as repair cost data, vehicle owner data, mileage data and damage event data, can be stored; that in a condition assessment process, the surface of the vehicle in a damaged area is optically captured in three dimensions using the optical detection unit, the spatial coordinate data of the damage area and from this the area coordinate data of the spatial coordinate data of the damage area are determined, from the spatial coordinate data of the damage area and the standard spatial coordinate data of the damage area the deformation depth data are determined and the area coordinate data and the associated deformation depth data are stored as damage data in a process data record;that in an evaluation and assignment process, the evaluation unit performs a database query for existing process data records with such existing area coordinate data that correspond to the area coordinate data of the damage area and, if existing process data records with corresponding area coordinates are found, these or parts thereof are output, whereby optionally the output of existing process data records is ordered according to the degree of correspondence between the deformation depth data determined in the damage area and the deformation depth data stored in existing process data records.

The application provides an electrocardiosignal enhancement and classification method based on a gated recurrent diffusion model, comprising the following steps: step 1, obtaining an electrocardiosignal dataset; step 2, setting parameters and hyperparameters of the diffusion model; step 3, establishing a neural network structure of the diffusion model; step 4, training the diffusion model, and supplementing the electrocardiosignal dataset by using the trained diffusion model; step 5, setting hyperparameters of a classification model; step 6, establishing the classification model; and step 7, training the classification model by using the electrocardiosignal dataset obtained in step 4, and using the trained classification model for electrocardiosignal classification. The method introduces the gated recurrent diffusion model to supplement data, compared with a traditional generative model, can capture time sequence information of the electrocardiosignal, and thus simulate the characteristics of the heart rate.

The embodiment of the invention relates to the technical field of voice recognition, and discloses a voice recognition method and electronic equipment, and the voice recognition method comprises the steps: carrying out the partitioning of target voice data based on a partitioning time length, so as to obtain a plurality of voice segments, and enabling the plurality of voice segments to be used for carrying out the voice recognition processing according to a time sequence; generating supplementary data, wherein the supplementary data is non-acoustic data; and performing voice recognition processing on the plurality of voice segments and supplementary data to obtain a voice recognition result, the supplementary data being located behind the voice segment at the tail of the time sequence. According to the invention, the integrity and accuracy of speech recognition can be improved.

The application provides a trackless navigation texture information redundancy analysis method, comprising: processing original image data of a current wear area by using an edge detection algorithm to obtain wear boundary contour coordinates and gray distribution data; extracting boundary coordinate point sets from a reliable anchor point area and determining an optimized update frequency value according to the position and shape matching degree of the boundary coordinate point sets and historical path map data; incrementally updating the path map by using the optimized update frequency value, integrating the gray distribution data of the reliable anchor point area and the texture detail loss degree value, and generating an enhanced positioning reference data set; updating the enhanced positioning reference data set according to supplementary data and performing periodic matching degree evaluation according to the optimized update frequency value to maintain long-term positioning consistency.