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

34results about How to "Increase data dimension" patented technology

Surrounding type hybrid optical field imaging device and method

The invention provides a surrounding type hybrid optical field imaging device and method. The method includes parameter calibration of a surrounding type hybrid optical field imaging device, and also includes using the surrounding type hybrid optical field imaging device to obtain images, fusing the obtained images, rendering the obtained images, and generating a final image. The surrounding type hybrid optical field imaging device and method provided by the invention combine a parameter calibration means with the depth capability of an optical field camera and a high-definition image of a traditional camera, and can obtain a high-quality scene three-dimensional image.
Owner:丁志宇

Method and system for determining fault characteristics of active power distribution network

The invention discloses a method and a system for determining fault characteristics of an active power distribution network. The determination method comprises the steps of S1, generating sample databased on acquired historical operation data; S2, performing feature selection on the sample data by using a support vector machine model-recursive feature elimination algorithm, and extracting fault features; S3, substituting the data corresponding to the fault features in the sample data into a fault prediction model built based on a support vector machine, determining the fault prediction accuracy of the fault prediction model, setting the fault feature as the fault feature of the active power distribution network when the fault prediction accuracy meets the requirement, or otherwise, executing the step S1 to regenerate the sample data. According to the fault feature determination method provided by the invention, the prediction efficiency of the fault prediction model is improved, a reference is provided for power distribution network operation and maintenance personnel, and the labor intensity of the maintenance personnel is reduced.
Owner:CHINA ELECTRIC POWER RES INST +2

System and method for realizing intelligent early warning of fraudulent transactions

The invention relates to a system and a method for realizing intelligent early warning of fraudulent transactions, the system comprises a protocol processing layer, a capability opening layer, a security control layer, a scene combination layer, an operation management layer and a network management layer, and the method comprises the following steps: fusing multi-source heterogeneous data; performing complex multi-source behavior feature modeling based on deep learning; performing anti-fraud intelligent identification based on incremental learning; constructing and analyzing a fraud behaviorobject association network map based on deep learning of the graph structure; constructing and reasoning an anti-fraud rule knowledge graph; performing decision engine construction. The technical problems that linkage between data sources of an existing anti-fraud transaction technology model is insufficient, and the fraud rule feature networked optimization requirement cannot be met can be solved.
Owner:上海欣方智能系统有限公司 +1

Micro-seismic signal multi-scale denoising method and device and readable storage medium

The invention discloses a micro-seismic signal multi-scale denoising method and device and a readable storage medium, and the method comprises the steps: 1, obtaining a micro-seismic signal, carryingout the EMD or EEMD decomposition, and filtering the high-frequency noise in the decomposed signal; 2, respectively constructing a Hankel matrix of each IMF component; 3, carrying out singular value decomposition is based on the Hankel matrix of each IMF component to obtain a score vector of principal component analysis, carrying out primary denoising, and carrying out soft threshold secondary denoising on each component signal and residual component after primary denoising; and 4, superposing the component signal subjected to secondary denoising and the residual component to obtain a denoisedmicro-seismic signal. According to the method, singular value decomposition is associated with principal component analysis, information of singular value decomposition serves as a score vector of principal component analysis, the PCA calculation process is simplified, and the defect that denoising cannot be conducted on a single column vector through the score vector of singular value decomposition is overcome.
Owner:CENT SOUTH UNIV +1

Method and device for loan customer credit scoring

Embodiments of the invention disclose a method and a device for loan customer credit scoring, and relates to the field of data processing. The method and the device can improve data dimension of credit scoring model analysis and accuracy of credit scoring results of loan customers. The method includes: obtaining at least one category of customer information of a sample customer, wherein the at least one category of customer information is associated with a unique identification code of the sample customer; preprocessing the customer information of each category to obtain sample data; accordingto the sample data corresponding to the customer information of each category and default information of the sample customer, modeling to generate a scoring model corresponding to the customer information of each category; obtaining a credit scoring model by fusing the scoring models corresponding to the customer information of each category; inputting the customer information of the customer tobe evaluated into the credit scoring model, and calculating the default information of the customer to be evaluated.
Owner:BANK OF CHINA

Attribute reasoning and product recommendation method based on adaptive graph convolutional network

The invention discloses an attribute reasoning and product recommendation method based on an adaptive graph convolutional network. The method comprises the following steps: 1, constructing heterogeneous data: a scoring matrix of a user to a product, a user attribute matrix, a product attribute matrix, a user attribute index matrix and a product attribute index matrix; 2, performing missing value filling preprocessing on the user attribute matrix and the product attribute matrix; 3, obtaining a cooperative matrix through one-hot coding; 4, constructing a feature fusion layer according to the attribute matrix and the collaborative matrix; 5, carrying out feature propagation through a graph convolution layer; 6, constructing a prediction layer to perform attribute reasoning and product recommendation; 7, updating the node attribute matrix according to the output result of the prediction layer; and 8, repeating the steps 4-7 until the attribute reasoning and product recommendation effectsare optimal. According to the method, the high-order structure information of the graph, the internal interaction between the node attributes and the potential association between the node attributesand the link relationship can be fully mined, so that more accurate attribute reasoning and product recommendation are realized.
Owner:HEFEI UNIV OF TECH

Speech recognition method, electronic equipment and storage device

The invention discloses a speech recognition method, electronic equipment and a storage device. The method comprises the following steps: acquiring to-be-recognized data when a user speaks, wherein the to-be-recognized data comprises audio data and video data of the mouth of the user; extracting a first feature representation by using the video data, and extracting a second feature representation by using the audio data; performing the following identification steps on the to-be-recognized data for several times: acquiring a fused context representation of the video data and the audio data by using the first feature representation, the second feature representation and the predicted text identified last time, and performing prediction by using the fused context representation to obtain the predicted text identified this time; and taking a combination of the predicted characters identified for several times as a final identification text of the to-be-recognized data. According to the scheme, the accuracy of speech recognition can be improved.
Owner:IFLYTEK CO LTD

Portfolio sector allocation method and device, server, and storage medium

ActiveCN107527289AImplement sentiment metricsScientific and efficient configurationFinanceWeb data indexingThe InternetSearch terms
The embodiments of the invention disclose a portfolio sector allocation method and device, a server, and a storage medium. The portfolio sector allocation method includes the following steps: capturing financial text data on the Internet; acquiring hot search terms related to stock market concepts and the search volume thereof from log data of a search engine; extracting at least one stock name associated with the hot search terms from the financial text data, and calculating the correlation intensity value between the hot search terms and each related stock; and calculating the decision parameters for deciding the sectors to which the at least one stock name belongs according to the search volume of the hot search terms and the correlation intensity values so as to carry out sector allocation according to the decision parameters. According to the embodiments of the invention, text data on the Internet and in the search engine are analytically processed using technologies such as big data mining and natural language processing, so as to realize reasonable sector allocation. The data dimension is expanded, and the decision-making process is quantified and more scientific and efficient.
Owner:SHANGHAI YOUYANG XINMEI INFORMATION TECH CO LTD

Method and system for automatically acquiring data model

The invention relates to the technical field of internet, and discloses method and system for automatically acquiring a data model. The method comprises the steps of configuring a data access rule; defining and building the data model according to the type and demand of acquired data; analyzing the data according to the forms and rules of the acquired data; filtering the analyzed data; mapping the filtered data to the data model again for matching; configuring a display module to the successfully matched data; starting to acquire. Compared with the prior art, the method has the advantages that the real-time big data stream technology is adopted, the advantage of big message data is fully utilized, and the effective information are mined to provide support to a target; the accuracy of the complementing result is increased as the dimension of the data adopted increases.
Owner:北京中交兴路车联网科技有限公司

Financial analysis report generation method and device

ActiveCN111061704AFit stateFit the development trendDigital data information retrievalFinanceData authenticityBusiness enterprise
The invention discloses a financial analysis report generation method and a device. The method comprises the steps of obtaining a data message of a target object; performing data arrangement on the data message by using the structured query statement to generate a standard message, and performing data cleaning; carrying out field extraction and calculation on the cleaned standard message accordingto an index construction rule to generate a financial index; and calculating the deviation rates of the same financial indexes from different sources, counting the financial indexes and the deviationrates, generating a financial analysis report, and outputting the financial analysis report. By applying the method of the invention, the cross validation among various data can be carried out by utilizing the deviation rate, so that the analysis direction of financial analysis, namely the analysis of the authenticity of financial data, is increased, meanwhile, the quality of data used later is ensured, and the data dimension and depth are expanded. And the financial analysis result is more accurate and comprehensive, so that the report can better conform to the overall state and developmenttrend of an enterprise.
Owner:东方微银科技股份有限公司

Indoor positioning method based on WiFi hot spot

The invention discloses an indoor positioning method based on a WiFi hot spot, which comprises an offline stage and an online stage, wherein the offline stage comprises the steps that a sample reference point and an AP node for receiving the WiFi hot spot are set in a designated area range, and information of all the AP nodes of all the reference sample points is collected; the mapping relation between the information of each AP node collected by each sample reference point and the position information of the corresponding reference point is calculated, and a fingerprint database is established; and a characteristic fingerprint space of the fingerprint database is extracted by adopting a PCA analysis method based on MapReduce. The online stage comprises the following steps: an RSSI signalof each AP received at each sample reference point is collected on line through a positioning terminal, and therefore an RSSI signal vector is formed; the PCA analysis method is adopted to extract themain positioning characteristic of the RSSI signal vector, so as to obtain a characteristic vector of the RSSI signal of each sample reference point; and accurate positioning is carried out accordingto the characteristic vector of the RSSI signal and the data information of the characteristic fingerprint space by adopting an EWKNN algorithm based on MapReduce.
Owner:NANJING UNIV OF POSTS & TELECOMM

Collaborative filtering recommendation method based on enhanced graph learning

The invention discloses a collaborative filtering recommendation method based on enhanced graph learning, and the method comprises the steps: 1, constructing a bipartite graph of a user to a product, wherein the bipartite graph comprises a user node set, a product node set and an adjacent matrix; 2, obtaining an embedded matrix as a node feature through one-hot coding; 3, calculating a similar matrix according to the current node characteristics, and performing sparsification; 4, taking the sparse similar matrix as a residual item and adding the residual item to the adjacent matrix to obtain an enhanced adjacent matrix; 5, carrying out feature propagation according to the enhanced adjacent matrix structure graph convolution layer, and obtaining node representation; and 6, obtaining a score matrix from the prediction layer according to the node representation, thereby realizing product recommendation. According to the method, the graph structure information can be adaptively learned based on the similarity between the nodes, and the robustness and integrity of the graph are enhanced, so that more accurate node representation is learned, and the recommendation performance is improved.
Owner:HEFEI UNIV OF TECH

Equipment fault prediction method and device, readable storage medium and electronic equipment

The invention discloses an equipment fault prediction method and device, a computer readable storage medium and electronic equipment. The method comprises the following steps: acquiring a fault sample data set of each reference equipment; determining a local fault prediction model corresponding to each fault sample data set; determining a target fault prediction model according to each local fault prediction model; according to the target fault prediction model, fault information of target equipment is predicted, and the similarity between the target equipment and the reference equipment meets a prediction condition. According to the technical scheme of the invention, the obtained target fault prediction model comprehensively considers the fault sample data sets of a plurality of reference devices, the data dimension is improved, and the data volume is increased, so that the fault information of the target device can be predicted more accurately.
Owner:ENNEW DIGITAL TECH CO LTD

A method and system for traceability of customs samples based on knowledge graph

The present invention provides a customs sample traceability method and system based on a knowledge graph, wherein the method includes: obtaining a first traceability code set and a first product image set according to the first product set to be inspected; Sorting the first source code set and the first product image set to obtain the first sorting result; based on the first sorting result, input the first source code set into the knowledge graph database for retrieval to obtain the first retrieval result; the attribute information of the first product Perform cluster analysis to obtain the first feature clustering result; build the first anomaly detection model based on the first feature clustering result; input the first associated product information and the first product image set into the first anomaly detection model to obtain the first Abnormal detection results, wherein the first abnormal detection result includes detecting abnormal products; generating a first recall instruction to recall the detected abnormal products.
Owner:智器云南京信息科技有限公司

A method and device for single-phase ground fault line selection in small current systems based on distribution automation data

The invention discloses a distribution automation data-based single-phase ground fault line selecting method and device for a small-current system. The distribution automation data-based single-phaseground fault line selecting method comprises the following steps: acquiring distribution automation data, and extracting the fault characteristic quantity of each distribution automation terminal; according to the fault characteristic quantity of each distribution automation terminal, constructing a fault measure membership function and a weight coefficient function of each distribution automationterminal; based on the fault measure membership function and the weight coefficient function of each distribution automation terminal, constructing a line fault membership function; calculating optimal values of line fault membership function parameters by a genetic algorithm, calculating the line fault membership of each line, judging whether a line has a fault according to the line fault membership of each line, and selecting a fault line. The distribution automation data-based single-phase ground fault line selecting method and device improve the fault line selecting accuracy, are applicable to a neutral ungrounded system and an arc suppression coil grounded system, and can identify various grounding conditions such as high-resistance grounding, arc grounding, intermittent grounding and instantaneous grounding.
Owner:DEZHOU POWER SUPPLY COMPANY OF STATE GRID SHANDONG ELECTRIC POWER

Domain name detection method and device, equipment and storage medium

The invention provides a domain name detection method and device, equipment and a storage medium, and relates to the technical field of network security. The method comprises the following steps: acquiring a domain name to be detected and inter-domain information corresponding to the domain name to be detected; associating the domain name to be detected with inter-domain information corresponding to the domain name to be detected; and inputting the associated domain name to be detected and the inter-domain information corresponding to the domain name to be detected into a pre-trained domain name classifier to obtain a detection result of whether the domain name to be detected is the target domain name or not.
Owner:CHINA TELECOM CORP LTD

A method for constructing a rapid quantitative detection model of heavy metal element copper in mulberry leaves

The invention provides a method for constructing a rapid quantitative detection model of copper, a heavy metal element in mulberry leaves, and belongs to the technical field of quality and safety detection. The invention first constructs a quantitative detection model of heavy metal copper, and then performs quantitative detection of heavy metal copper in blind mulberry leaf samples according to the quantitative detection model. When constructing the quantitative detection model, the present invention performs non-supervised clustering of the self-organizing neural network on the laser-induced breakdown spectrum data, and then uses the non-information variable elimination method to select the clustered variables, thereby avoiding information redundancy , to obtain the most relevant model variable information, and then establish a partial least squares regression model, select a partial least squares regression sub-model with high accuracy and stability, and fuse the partial least squares regression sub-model through a consensus fusion algorithm, A quantitative detection model for heavy metal copper was obtained.
Owner:ZHEJIANG UNIV

Method for managing traffic item in software-defined networking

A method for managing traffic item in software defined networking includes establishing a downlink flow table of a switch according to the flow entries, establishing an uplink flow table of the switch according to the flow entries, acquiring a data packet by the switch, and generating a transmission path to allocate the data packet according to the data packet, the downlink flow table, and the uplink flow table. The downlink flow table includes a correlation between first transmission ports of the switch and down link switches. The uplink flow table includes a correlation between the first transmission ports and a transmission port group of uplink switches.
Owner:INVENTEC PUDONG TECH CORPOARTION +1

Genetic diagnosis system integrating data acquisition, analysis and report generation

PendingCN113470776AAvoid repeatedly asking for patient informationAvoid the problem of hindering information acquisition caused by patients' inconvenient communicationProteomicsGenomicsMedicineData mining
The invention relates to a genetic diagnosis system integrating data acquisition, analysis and report generation. The system comprises a clinical module, a medical history module, an association module and a report module. The medical history module carries medical history data of a patient and a plurality of groups of family data distinguished based on family tags, wherein a plurality of identity tags are linked to the family tags, and every two identity tags are associated through family attributes; the association module is respectively connected with the clinical module and the medical history module, and associates the clinical data with the medical history data of the patient through the identity tags, so that the identity tag of the patient can be taken as a center, and a family data network can be formed through outward expansion of family attributes; and the personal clinical data and the medical history data of the corresponding members are linked through the identity tags of the members of different families, so that a multi-dimensional family data network is formed, and comprehensive and accurate genetic diagnosis is realized.
Owner:DERMATOLOGY HOSPITAL SOUTHERN MEDICAL UNIV (GUANGDONG PROVINCIAL DERMATOLOGY HOSPITAL GUANGDONG PROVINCIAL CENT FOR STI & SKIN DISEASES CONTROL & PREVENTION RES CENT FOR LEPROSY CONTROL & PREVENTION CHINA)

Online retrieval method and system

The invention discloses an online retrieval method and system. The method comprises the steps of constructing query points based on query operation; searching a neighbor point of which the direction is closest to the query point in a direction navigation map; and determining at least one return point closest to the query point based on the operation of searching the neighbor point closest to the query point in the direction. According to the method, the concept of the direction navigation graph is introduced on the basis of the direction vector, and the attribute that the direction navigation graph can be used for performing direction pruning on neighbors is utilized, so that the graph retrieval efficiency is greatly improved, and the method can adapt to online retrieval application scenes with larger data scales and higher data dimensions.
Owner:ALIBABA GRP HLDG LTD

Atmospheric Dust Pollution Baseline Stripping Method Based on Distributed Algorithm

ActiveCN107402171BCalculation method is reasonableGet rid of the one-sided disadvantage of single-point detection resultsParticle suspension analysisSpecial data processing applicationsAtmospheric dustSensor node
The invention relates to an atmospheric fugitive dust pollution baseline removal method based on a distributed algorithm. The method comprises the following steps: fugitive dust concentration sensor nodes are distributed according to the area of a monitoring zone; values of fugitive dust concentration in the monitoring zone are acquired and transmitted to a cloud server; the cloud server receives the sensor data, takes the sensor data as raw data and calculates a fugitive dust pollution baseline of the monitoring zone with a multi-dimensional data fusion algorithm; the baseline in the raw data is removed by means of the fugitive dust pollution baseline of the monitoring zone, and fugitive dust discharge amount corresponding to the environmental pollution baseline in a specific time interval is obtained; the total value of the fugitive dust discharge amount in the same time interval is calculated on the basis of the raw data; the environmental pollution baseline discharge amount is removed on the basis of the total value of the fugitive dust discharge amount, and human factor discharge amount of the monitoring zone is obtained. With adoption of the method, an enterprise is only responsible for self-produced fugitive dust, and the fugitive dust calculation way is more reasonable; the problem that a detection result of single fugitive dust point is one-sided is solved, and the core of the problem can be solved directly by means of real-time remote monitoring.
Owner:NORTHEASTERN UNIV LIAONING

Multi-target human body activity sensing device based on radar echo analysis algorithm

The invention discloses a multi-target human body activity sensing device based on a radar echo analysis algorithm, which comprises a preceding-stage radar module and a data processing and main control unit, and is characterized in that the preceding-stage radar module is used for receiving and sending signal data and performing data transmission with the data processing and main control unit; the data processing and main control unit comprises a main control MCU, the main control MCU comprises a waveform calibration module and a data processing module, the waveform calibration module carries out echo processing analysis and identification and sends processed data to the data processing module, and the data processing module completes identification of waveform basic feature points and carries out data processing. The novel sensing device is used in an indoor space, can identify and comprehensively measure the multi-target activity behavior amplitude according to a radar echo analysis principle, and normalizes the multi-dimensional data to measure the human body thermal environment comfort level load index.
Owner:周旭

Mechanical ventilation man-machine asynchronous detection method and device based on graph neural network

The invention discloses a mechanical ventilation man-machine asynchronous detection method and device based on a graph neural network, a storage medium and equipment. The detection method comprises the following steps: acquiring real-time waveform data of a to-be-detected object in a mechanical ventilation process; converting the obtained real-time waveform data into real-time graph structure data; and inputting the real-time graph structure data into a pre-trained graph convolutional neural network model to obtain a man-machine asynchronous type. According to the detection method, the waveform data is converted into the graph structure data in the non-Euclidean space, and the graph convolutional neural network model is adopted to predict the man-machine asynchronous type, so that the data dimension is improved, the data analysis freedom degree is enhanced, and the visual sense of analysis processing is widened; therefore, the man-machine asynchronous type prediction accuracy of the to-be-tested object in the mechanical ventilation process is improved.
Owner:SHENZHEN INST OF ADVANCED TECH

Method and device for monitoring grain condition in bulk grain container

The invention discloses a method and device for monitoring grain condition in a bulk grain container, and the method comprises the steps: collecting temperature and humidity data, obtaining the temperature and humidity data of the environment where grains are located in real time, transmitting the temperature and humidity data to a collection terminal, marking a time label on the received temperature and humidity data by the collection terminal to generate timing temperature and humidity data; transmitting the timing temperature and humidity data to a remote server; performing temperature and humidity data processing: enabling the remote server to perform filtering and wavelet analysis on the timing temperature and humidity data, interpolate and eliminate repeated data, and obtain continuous temperature and humidity data of each acquisition node; and performing temperature and humidity data fusion: fusing the continuous temperature and humidity data by using a k-medoids clustering algorithm, carrying out fusion once in each acquisition period, fusing the continuous temperature and humidity data of the plurality of acquisition nodes into a piece of continuous overall temperature and humidity data, and reflecting the overall state of the environment in which the grains are located by the overall temperature and humidity data. The invention further discloses a monitoring device applying the monitoring method.
Owner:CHINESE ACAD OF AGRI MECHANIZATION SCI
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products