36 results about "Data discrimination" patented technology

The invention relates to the technical field of data acquisition and processing, and provides a real-time data acquisition and processing method and system for a distributed control system, and the method comprises the steps: collecting a data flow in real time through a terminal sensor, analyzing the probability distribution of a data change rate through a sliding window, and obtaining a data change rate; and a sensitivity threshold is intelligently generated by combining dynamic factors such as equipment electric quantity and network load. When data mutation exceeds a threshold value, the system automatically extracts a key change section and generates time, period and variable quantity three-dimensional features, and meanwhile, the frequency domain analysis result is fused to enhance the anti-interference performance. And the event emergency degree is calculated based on the time-space correlation of the spectrum energy entropy and the characteristic parameters, and finally a processing queue sorted according to priorities is formed and the bandwidth is dynamically allocated. According to the method, through multi-dimensional data fusion and resource adaptive adjustment, the response speed of the system to emergencies, the data discrimination precision and the stability in a complex environment are improved.

To solve problems in utilizing data.SOLUTION: An automatic data discrimination and proposal AI system for solving problems in utilizing data includes means for classifying data, means for extracting useful information, and means for predicting missing data and generating utilizing ideas.SELECTED DRAWING: Figure 1

The invention relates to the technical field of data processing and machine learning, and discloses a multi-modal clustering method, device and equipment, a storage medium and a product, and the method comprises the steps: determining the specific representation of original multi-modal data based on an auto-encoder and reconstruction loss, and carrying out the reconstruction of the original multi-modal data based on the specific representation; obtaining a clustering prototype through an adaptive average pooling operation and an expectation maximization algorithm, generating common characterization of the original multi-modal data according to a cross attention mechanism and the clustering prototype, carrying out comparative learning on the specific characterization and the common characterization to obtain aligned common characterization, and clustering the aligned common characterization to obtain a clustering result; and obtaining a clustering label of the original multi-modal data. According to the method, the unique representation is extracted through the self-encoder, the representation of the clustering prototype is dynamically adjusted by adopting the self-adaptive average pooling operation and the expectation maximization algorithm, and the common representation is generated based on the cross attention mechanism and the clustering prototype, so that the clustering label is obtained, the model data distinguishing capability is enhanced, and the accuracy of the clustering label is improved.

This invention belongs to the field of target detection and recognition technology, and proposes a quantum radar experimental device based on a non-classical optical field. The device comprises a transmitter, a receiver, and a transmission-receiver synchronization system. The transmitter includes a light source unit, a non-classical optical field control unit, and a signal collimation and transmission unit. The receiver includes a signal collection unit, a detection unit, and a data discrimination unit. The transmission-receiver synchronization system enables clock synchronization and real-time communication between the transmitter and receiver. This invention can effectively overcome the high dependence of existing quantum radars on quantum entanglement, improve the ease of operation and equipment of quantum radar, enhance the detection capability of dark and small targets, and provide key technical support for the large-scale application of quantum radar in the future.

The present application discloses a method and device for fusing space weather observation and simulation based on a 3D adversarial network, relating to the technical field. The method includes: interacting with a space weather simulation database, and obtaining simulation data based on a preset selection constraint; performing adaptive grid division on the simulation data based on adaptive grid technology to obtain divided adaptive grids, and storing the divided adaptive grids in an associated manner with the simulation data; obtaining observation data, and performing grid matching of the observation data based on the divided adaptive grids and the simulation data; inputting the observation data and the simulation data after grid matching into a 3D GAN generator to obtain generated data, wherein the generated data is marked with gradient information; activating a 3D GAN discriminator, performing data discrimination based on the generated data, and outputting fused data based on the data discrimination result. Thus, the technical effects of improving the data fusion accuracy and enhancing the fusion adaptability are achieved.

To provide a game machine that allows a processing load to be reduced.SOLUTION: Control data for executing predetermined control on a game is stored in control data storage means. Control of the game is executed by control means according to the control data stored in the control data storage means. On the basis of establishment of a predetermined storage condition, the control data is stored in the control data storage means by storage control means. On the basis of establishment of the predetermined storage condition, at least one of control data stored in the control data storage means and data stored in the control data storage means is discriminated by data discrimination means. On the basis of a fact that a discrimination result of the data discrimination means becomes a specific discrimination result determined beforehand. Control is executed so that data according to the control data is stored only in a part of a storage region of the control data storage means. Thereby, an effect of reducing a processing load is produced.SELECTED DRAWING: Figure 1

The invention relates to a data retrieval mining processing method and system for a character database, and relates to the technical field of data processing, and the method comprises the steps: obtaining a target character and an event category, inputting the categories into a database, and retrieving to obtain associated data, types and update time; analyzing the priorities of the data types and arranging the data types into a first data sequence; combining the target event category related time with the data updating time, adjusting the type priorities, obtaining the adjusted type priorities, and arranging the adjusted type priorities into a second sequence; and generating a basic weight and a time weight according to the target event category, optimizing the two data sequences to obtain an optimal display data sequence, and displaying the optimal display data sequence as a retrieval mining result. According to the method and the device, the problems that the traditional character data retrieval display is only ordered according to time correlation or types, the applicability is insufficient, and the data distinguishing and calling efficiency and accuracy are influenced are solved, and the comprehensive applicability and retrieval accuracy of data display are improved.

The embodiments of the present disclosure disclose a model training method, a data discrimination method, equipment, a medium and a product. The method comprises: obtaining time series data; generating synthetic time series data through a time series data generation model; obtaining time series feature data, obtaining synthetic time series feature data; obtaining a data discrimination result; training the time series data generation model and the time series data discrimination model through an adversarial training method based on the data discrimination result; and determining the time series data generation model after at least one round of training as a target time series data generation model in response to the data discrimination result output by the time series data discrimination model after at least one round of training satisfying a synthetic time series data indistinguishable condition. The target time series data generation model determined based on the scheme can generate synthetic time series data with high similarity to real time series data.

The invention discloses a sensor attack detection method and device in the Internet of Vehicles, a medium and a product, and relates to the field of intelligent vehicles, and the method comprises the steps: carrying out the preprocessing of the data of a real vehicle; according to the preprocessed data, a novel generative adversarial network is adopted to obtain generated data; the novel generative adversarial network comprises an abnormal data generation model and an abnormal data discrimination model based on an adaptive self-attention mechanism; performing attack detection according to sample data composed of the generated data and the preprocessed data and the preprocessed data; according to a detection result, a dynamic adjustment module based on an adaptive feedback mechanism is adopted to update the novel generative adversarial network; generating abnormal data by using the updated novel generative adversarial network; and performing sensor attack detection according to the generated abnormal data. According to the invention, accurate, efficient and real-time optimized Internet of Vehicles data analysis and network attack detection can be realized, so that the safe operation of the intelligent driving vehicle is ensured.

The embodiment of the invention discloses a health assessment method and device, electronic equipment and a computer readable storage medium. The method comprises the following steps: acquiring physiological data of a target object; the method further comprises the step of determining physiological discrimination data of the target object based on the physiological data of the target object and physiological data discrimination conditions, and the physiological data discrimination conditions are dynamically adjusted at least based on historical data of similar objects. The method further comprises the step of determining historical doctor seeing data of the target object based on historical data of the target object. In addition, the method further comprises the step of outputting a health assessment result of the target object by the assessment model based on the physiological discrimination data and the historical treatment data of the target object. Through the mode, the physiological data detected by the target object in real time is judged based on the dynamically adjusted judgment condition, the physiological judgment result of similar objects can be accurately reflected, the physiological judgment data of the target object and the historical treatment data are combined, and the personalized health state of the target object is accurately evaluated.

The invention provides a pressure change memory. Each pressure change memory unit comprises a first electrode, a first dielectric layer, a memory function layer, a second dielectric layer and a second electrode which are sequentially arranged from top to bottom, the storage function layer has an amorphous characteristic and a defect state with the concentration higher than 10 < 14 > cm <-3 >, and the component is selected from a sulfur threshold switch material, an oxide, a nitride or a solar cell material. The device realizes controllable offset of threshold voltage under the action of forward and reverse electric fields, and takes the voltage difference as a read-write window, thereby realizing nonvolatile information storage. Through asymmetric structure design, storage function layer multi-layer design or PN junction introduction, a read-write window can be further expanded, and the reliability of data discrimination is improved. The memory is simple in structure and easy to integrate, has the characteristics of high speed and low power consumption, and is suitable for three-dimensional stacked high-density storage application.

The present application relates to the technical field of Internet advertising delivery, and in particular to an advertising delivery data anti-cheating method, system, terminal and medium, aiming to solve the problem in the prior art that traffic owners, in order to obtain higher profits, mix fictitious data into the advertising delivery data provided to advertisers. Its technical solution is an advertising delivery data anti-cheating method, including: obtaining an advertising delivery request package and extracting matching advertising data; creating a behavior data reporting link, combining it with the advertising data and sending it to the traffic main server; collecting and storing behavior data; screening and removing false data in the behavior data through a preset false data identification process, and obtaining valid behavior data after removing the false data; settling advertising service fees based on valid behavior data. The present application has the effect of reducing false data when settling advertising service fees and improving the correlation between advertising service fees and advertising delivery effects.

The invention belongs to the technical field of analysis methods, and particularly relates to an animal behavior analysis method and device based on a multi-modal agent, computer equipment and a storage medium. The method comprises the following steps: receiving multi-view video data and a natural language cue word; extracting three-dimensional skeleton key point data of the animal from the multi-view video data; the behavior analysis agent generates semantic description and a calculation task list at least comprising one physical index according to video key frames and natural language cues in the multi-view video data; calling an independent calculation tool chain, and performing quantitative calculation on the three-dimensional skeleton key point data according to the calculation task list to obtain a quantitative result corresponding to the physical index; and fusing the semantic description and the quantitative result through a slot filling mechanism to generate an animal behavior analysis report. According to the method, the physical accuracy of calculation can be realized, the capability of generating a scientific and interpretable analysis report is also realized, and the crossing from data discrimination to intelligent analysis is realized.

The present invention provides an information processing device that can perform processing using weight measurement data more appropriately. [Solution] The information processing device 30 comprises a measurement information acquisition unit 331, a data discrimination unit 332, a data processing execution unit 334, and a notification unit 335. The measurement information acquisition unit 331 performs an acquisition process to acquire weight measurement data from a measuring device installed in the pet toilet. The data discrimination unit 332 performs a pattern discrimination process to determine which of one or more patterns the weight measurement data group corresponds to. The data processing execution unit 334 and the notification unit 335 perform processing according to the pattern determined by the pattern discrimination process.

The invention relates to a method for improving LIBS (laser-induced breakdown spectroscopy) quantitative analysis precision based on abnormal data detection and removal, in particular to a method for obtaining spectral data based on a laser-induced breakdown spectroscopy (LIBS) technology, carrying out abnormal data discrimination on a spectrum before establishing a quantitative analysis model by using the data, and carrying out quantitative analysis on the spectrum. Homogeneous sample abnormal data and heterogeneous sample abnormal data in a calibration sample data set are searched, the abnormal data are removed, then training is carried out, a prediction model is established, and the prediction precision and robustness of the model are improved. Wherein the discrimination and detection of the abnormal data need to carry out principal component analysis (PCA) on calibration sample data. By utilizing the method, a model calibration error caused by abnormal spectral signals due to environment or human factors in an LIBS experiment process or a measurement error caused by introduction of different samples can be avoided, the predictive capability and robustness of an LIBS quantitative analysis model are improved, and high-precision application of LIBS equipment in various application scenes is met.

The invention discloses a data deduplication method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the steps of obtaining to-be-processed data, and determining first to-be-processed data corresponding to each interface according to the interface corresponding to each piece of data; for the first to-be-processed data of each interface, calculating the data distinction degree of each feature field, and determining a target feature field according to the data distinction degree; and performing feature data extraction through the target feature field, and performing data deduplication based on the feature data. According to the embodiment, the data deduplication accuracy can be greatly improved.

The application provides a data security monitoring method, device, intelligent gateway, system and storage medium for coal mine gas emission reduction. The method comprises the following steps: acquiring first time series data collected by at least one sensor of a to-be-monitored device within a recent preset time period; generating a dynamic threshold according to the first time series data; acquiring second time series data collected by the at least one sensor of the to-be-monitored device; detecting whether the second time series data is within the dynamic threshold range; in the case that the second time series data is not within the dynamic threshold range, marking the second time series data as abnormal data; otherwise, marking the second time series data as normal data; and transmitting the second time series data marked as normal data to a national carbon market management platform. The embodiment of the application can transfer the task of abnormal data discrimination to the intelligent gateway side, greatly reducing the data transmission amount of the gateway and the national carbon market management platform, and avoiding waste of network bandwidth and storage resources.

The present application discloses a sensor attack detection method, device, medium and product in the Internet of Vehicles, which relates to the field of intelligent vehicles. The method includes preprocessing the data of real vehicles; using a new generative adversarial network to obtain generated data based on the preprocessed data; the new generative adversarial network includes: an abnormal data generation model and an abnormal data discrimination model based on an adaptive self-attention mechanism; performing attack detection based on sample data composed of generated data and preprocessed data and preprocessed data; updating the new generative adversarial network based on a dynamic adjustment module based on an adaptive feedback mechanism according to the detection results; using the updated new generative adversarial network to generate abnormal data; and performing sensor attack detection based on the generated abnormal data. The present application can realize accurate, efficient and real-time optimized Internet of Vehicles data analysis and network attack detection, thereby ensuring the safe operation of intelligent driving vehicles.

The embodiment of the application discloses a kind of saliency target detection hybrid supervision training methods, by step, saliency target detection model of multiple different network structures is trained using labeled data image respectively, the discriminator of reliable pixel and unreliable pixel in unlabeled image is divided;Unlabeled data image and fine annotation data image are simultaneously loaded into the saliency target detection model to be trained as training data;For labeled data image, data augmentation method is used to increase the diversity of data samples, for unlabeled data image, the saliency prediction result corresponding to the image is obtained by calling the trained multiple discriminators respectively;Pixel reliability is divided;The reliable pixel in labeled data and unlabeled data is mixed supervision training to model, by data labeling, and by model corresponding data discrimination, the method for increasing the diversity of data samples mixes all data supervision training, improves the generalization ability of model.

The embodiment of the application discloses a method for obtaining the discrimination of financial service data, which comprises the following steps: obtaining a candidate financial service data set as a financial data sample, determining the first number of different values R1 of the financial data sample according to part of the data of the financial data sample, determining the second number of different values R2 of the financial data sample according to all the data of the financial data sample, and finally determining the discrimination of the financial service data according to the comparison relationship between the first number of different values R1 and the second number of different values R2, so that the related service data can be quickly inquired or retrieved according to the size of the discrimination of the data in the financial service data set.

The present invention discloses a storage management system for hazardous chemicals based on RFID automatic identification, which specifically relates to the field of ray identification technology, including an RFID information entry module for hazardous chemicals cabinets, a regional division module for hazardous chemicals cabinets, a cabinet data acquisition module, a cabinet data analysis module, a comprehensive data processing module, a comprehensive data discrimination module, and an early warning module. The present invention can more effectively perform more accurate detection in each sub-area through the regional division of hazardous chemicals cabinets, perform accurate comparison through comprehensive data discrimination, and divide the hazard index into hazard levels, making the discrimination more accurate and intuitive; through early warning, it can timely discover that the storage management system of hazardous chemicals cabinets sends out dangerous signals, and at the same time, the system can realize accurate management and real-time monitoring of chemical reagent bottles, quickly discover and handle problems, ensure the stability and safety of hazardous chemicals cabinets, and ensure the reasonable use and storage of chemicals.

The invention discloses a method for online verification of a natural gas measurement standard device, which relates to the field of natural gas measurement technology, and aims to solve the problem that the quality monitoring means can effectively detect the stability state of the accuracy of the measurement standard device, but inevitably there is hysteresis, and the performance change of the measurement standard cannot be grasped in the first time. The key points of the technical solution are as follows: S1: calculation of the center line; A1: calculation of the indication error; A2: calculation of the center line; S2: setting the control limit; S3: calculation of the early warning limit; B1: data discrimination and selection; B2: dynamic early warning calculation; B3: calculation of the control limit; B4: early warning limit of multiple standard tables; S4: over-limit judgment; C1: selection of early warning limit and control limit; C2: over-limit judgment of single standard table; C3: over-limit judgment of multiple standard tables. The effect of accurate measurement and calculation, measurement and calculation of multiple types of standard tables and convenient calculation of over-limit is achieved.

The invention discloses a light GBM-based unmanned aerial vehicle trajectory feature anomaly detection method, which overcomes the problems of low accuracy and high time complexity of unmanned aerial vehicle abnormal trajectory detection in the prior art, and comprises the steps of obtaining unmanned aerial vehicle flight data, and establishing an unmanned aerial vehicle flight data set; classifying the flight data according to the Euclidean distance between the data points by using a clustering algorithm, and removing take-off data and descent data; according to the processed flight data of the unmanned aerial vehicle, constructing a track feature vector of unmanned aerial vehicle cruising; and training an unmanned aerial vehicle trajectory anomaly detection model by using light GBM based on the trajectory feature vector, and carrying out unmanned aerial vehicle trajectory anomaly detection. According to the method, noise data in the flight data of the unmanned aerial vehicle is cleaned, data discrimination in different flight modes is carried out, data classification of different flight modes is completed, corresponding trajectory feature vectors are constructed, and the trajectory detection model is trained by using the lightGBM, so that the time complexity is lower, and the accuracy is higher.

The application provides a physical-data collaborative driving generated adversarial network (PIGAN) long-span bridge abnormal wind data self-adaptive completion system. The system comprises a long-span bridge abnormal wind data discrimination module 101, a long-span bridge abnormal wind data interpolation module 102 and a long-span bridge interpolated wind data evaluation module 103. The 101 utilizes a set priori criterion to discriminate the abnormal state of the measured wind data in real time, and completes abnormal marking and abnormal classification. The 102 interpolates the wind data of different abnormal types according to the abnormal classification result of the 101. If the abnormal wind data belongs to the accidental abnormal category, the weighted K-nearest neighbor (WKNN) algorithm is adopted. If the abnormal wind data belongs to the continuous abnormal category, the PIGAN is adopted. The 103 is used for evaluating whether the interpolated wind data reaches the preset quality requirement. If the preset quality requirement is not reached, the 102 is returned. If the preset quality requirement is reached, the long-span bridge abnormal wind data self-adaptive interpolation is completed.

The invention discloses an abnormal data discrimination method and device based on time series data analysis, and belongs to the field of time series data analysis. Sequentially calculating a neighborhood difference factor, a neighborhood difference factor time sequence, an initial segmentation point, a reference factor segment, a difference fluctuation degree, an initial smoothing parameter, a smoothing contribution degree, a smoothing weight and an adaptive smoothing parameter; and judging abnormal data according to the self-adaptive smoothing parameters. According to the method, the influence condition of multiple dimensions on data is better reflected by calculating the initial segmentation point, and the importance of the temperature and phosphate radical concentration in the reference factor segment on a subsequent abnormal data judgment result is better reflected by calculating the initial smoothing parameter; and an adaptive smoothing parameter is calculated by combining extreme difference conditions and continuous trend difference conditions in different reference factor segments and the influence degree of local data on the whole data segment, so that the accuracy of abnormal data discrimination is improved.

The invention relates to a data processing method and device based on multi-modal contrast fusion and a graph neural network, and a medium. The method comprises the following steps: collecting multi-modal biomedical data set samples of mRNA expression data, DNA methylation data and miRNA expression data; embedded matrixes of three modes are obtained through an independent feedforward neural network encoder; adopting a contrast learning mechanism based on NT-Xent to carry out unsupervised contrast alignment on the embedded matrixes of different modalities, and generating aligned embedded matrixes of three modalities; embedding and stacking the aligned modalities into a sequence, and inputting the sequence into a multi-layer Transform encoder to obtain a unified fusion representation; normalizing the fusion representation, and constructing a sample similarity graph structure; and based on the graph structure and the fusion representation, node high-order adjacency features are extracted through a three-layer graph convolutional network, sample classification is completed in combination with two full-connection layers, and a sample classification result is output. Compared with the prior art, the method has the advantages of cross-modal data fusion optimization, medical data discrimination enhancement, high efficiency and the like.

The invention discloses an intrusion data discrimination method and device, electronic equipment, a storage medium and a vehicle, and relates to the technical field of vehicle information security, and the method comprises the steps: extracting multi-dimensional operation data features in the collection of first vehicle operation data of the vehicle through a preset discrimination model, and constructing an operation data feature matrix; determining the confidence coefficient of the operation data feature matrix based on context information in the vehicle operation process; and according to second vehicle operation data of a plurality of vehicles of the same type, in combination with the confidence coefficient, discriminating the operation data features, and determining an intrusion data discrimination result. Multi-source data fusion analysis is realized, the characteristics of a vehicle as a mobile physical entity are fully considered, real attack and normal behavior variation are effectively distinguished, and the false alarm rate is reduced.

The application provides a multi-modal private data generation model training method, data generation method and system, and belongs to the technical field of data privacy protection.The application realizes the generation of RGB images and table data in a unified framework, does not need to use additional public data and an end-to-end training mode, and is connected with each other in the generation process of various types of data, thereby improving the semantic consistency of the generated data, effectively fusing the key information of different modalities to realize accurate multi-modal data discrimination;multi-angle discrimination and multi-modal matching loss are proposed, the multi-modal discrimination loss dominates the training process of the GAN model, the single-modal discrimination loss is used to assist the model in balancing the learning efficiency of different modalities, and the matching loss promotes the learning of the model on the relationship between modalities.In order to reduce the privacy leakage risk of the generation model, the differential privacy mechanism is applied to the image and table data joint generation model based on GAN, so as to balance the availability and privacy of the synthesized data.

The present invention discloses a vulnerability exploitation analysis method, system and storage medium based on big data, which relates to information security technology. A main information channel is constructed, and a data discrimination library is used to classify existing data into secure data and threat data according to a preset security level and complete the screening of the data; a base database is constructed based on feature point information, feature point threat coefficients, feature chains and feature chain threat coefficients; sorting is performed according to the numerical value of the threat coefficient C and the parsing order of new data is determined; a spider web line, an external boundary line, a standard data boundary map and a base map are constructed; based on the free quantity combination of unknown feature points and existing feature points on the spider web into unknown feature chains, a measurement surface and a comparison surface are constructed and the threat level of the data is initially judged. The present invention adopts a unified analysis algorithm, which improves the running speed and scalability of the algorithm while ensuring accuracy.

The invention discloses a polarization compass orientation method based on real-time discriminant analysis of interpretable course information, which comprises the steps of course data modeling, real-time monitoring, discriminant analysis and noise suppression, modeling by adopting a sparse local Fisher discriminant analysis method, enabling different types of course data states to be more separated according to an inter-class dispersion matrix of a weight coefficient matrix, and improving the orientation accuracy. Similar and same course data states are more clustered by using an intra-class matrix; a minimum absolute contraction selection operator optimization model is introduced, so that a course data discrimination model is simpler and more interpretable; course process data are analyzed through monitoring indexes, so that evolution and instability of dynamic output data of the polarization compass can be monitored in real time, and fine description of the data process can be realized; noise of discriminant analysis is effectively suppressed while real course signals in dynamic course data output by the polarization compass are reserved to the maximum extent. The method solves the problem that the orientation precision of the polarization compass is affected due to weak machine learning generalization ability.