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

62results about How to "Clustering implementation" patented technology

Clustering method and system of parallelized self-organizing mapping neural network based on graphic processing unit

The invention relates to a clustering method and system of a parallelized self-organizing mapping neural network based on a graphic processing unit. Compared with the traditional serialized clustering method, the invention can realize large-scale data clustering in a faster manner by parallelization of an algorithm and a parallel processing system of the graphic processing unit. The invention mainly relates to two aspects of contents: (1) firstly, designing the clustering method of the parallelized self-organizing mapping neural network according to the characteristic of high parallelized calculating capability of the graphic processing unit, wherein the method comprises the following steps of obtaining a word-frequency matrix by carrying out parallelized statistics on the word frequency of keywords in a document, calculating feature vectors of a text by parallelization to generate a feature matrix of data sets, and obtaining a cluster structure of massive data objects by the parallelized self-organizing mapping neural network; and (2) secondly, designing a parallelized text clustering system based on a CPU / GPU cooperation framework by utilizing the complementarity of the calculating capability between the graphic processing unit (GPU) and the central processing unit (CPU).
Owner:HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL

Highway road section traffic state discrimination method based on charging data

The invention relates to a highway road section traffic state discrimination method based on charging data. The method comprises the steps of: 1, dividing basic road sections of a highway; 2, processing highway charging data and obtaining a road section traffic volume curve; 3, comparing the similarity of the curves of all road section flows changing with time, and obtaining a similarity coefficient; 4, according to the road section similarity coefficient, dividing the road sections into categories of a preset number; and 5, discriminating the traffic states of each kind of road sections after the classification. By adopting the method, the traffic states of different road sections of the highway can be effectively judged.
Owner:BEIHANG UNIV

Method for detecting abnormal time sequence without class label

ActiveCN104899327AEnhanced couplingSegmentation results are compactRelational databasesCharacter and pattern recognitionSatellite dataNearest neighbour algorithm
The invention provides a method for detecting an abnormal time sequence without a class label, and aims at solving the problems that ideal effect of segmenting fixed points based on satellite remote detecting data cannot be achieved, the clustering number is manually set during layer-based clustering, and offline and online abnormality detection methods for the label time sequence without the class label are currently not developed. According to the technical scheme, the method comprises the steps of 1, segmenting the satellite remote detecting historical data according to the cycle property of the satellite remote detecting data to obtain the time sequence without class label, namely, X={x1, x2..., xn}; 2, performing adaptive layer-based clustering for the X={x1, x2..., xn} obtained in step 1, and determining and deleting the abnormal sequence in the time sequence without the class label to obtain the formulas as shown in the specification; 3, adopting the formulas as shown in the specification as samples, performing mode matching for the formula shown in the specification by the nearest neighbor algorithm according to the matching threshold, so as to finish the abnormal satellite remote detecting data detection. The method is applied to the field of satellite data detection.
Owner:HARBIN INST OF TECH

Satellite-ground joint beam forming and power distribution method based on non-orthogonal multiple access technology

The invention discloses a satellite-ground joint beam forming and power distribution method based on a non-orthogonal multiple access technology (NOMA), and the method comprises the steps that a satellite communication sub-network serves a plurality of ground stations, a ground cellular sub-network serves a plurality of cellular users, and two sub-networks achieve the spectrum sharing; firstly, anNOMA technology is adopted in the ground cellular sub-network, and a multi-user clustering method is provided based on the correlation of channels among cellular users and the difference of channel gains; secondly, under the conditions that the user service quality is guaranteed and the transmitting power of a satellite and a base station is limited, the optimization problem of sum rate maximization of the whole system is established; and, the non-convex problem is converted into a convex optimization problem by adopting an S-Procedure and Taylor expansion method, and further the optimal beamforming weight vector and the power distribution factor of the satellite and the base station are obtained by utilizing an iterative penalty function method. According to the method, the beam formingtechnology and the non-orthogonal multiple access technology are combined, and a technical scheme is provided for improving the effectiveness of satellite-ground fusion network information transmission.
Owner:NANJING UNIV OF POSTS & TELECOMM

Overall reconstruction design method of plane line position of existing railway

The invention discloses an overall reconstruction design method of a plane line position of an existing railway. The method comprises the steps that the line element types of testing points are identified based on the tangent azimuth change rate of each testing point, and initial clustering of the testing points is conducted; based on the number of the testing points in each line element point group, the initial clustered line element point groups are adjusted; based on the crossing point position of straight lines at two ends of each circular curved segment, the line element point groups arefurther adjusted; iterative computation is conducted, easement curve line element testing points in a linear element point group and a circular curve line element point group are gradually identifiedand adjusted, so that the number of the testing points in the line element point groups stable, final clustering of three kinds of line element testing points is achieved, and fitting of local line positions is conducted; all the local line positions are connected and form an initial overall fitting line position, the fitting line position is optimized, and the final plane reconstruction scheme ofthe existing railway is obtained. The overall reconstruction design method of the plane line position of the existing railway can precisely identify different types of line element testing points, and can optimize the fitting line position from an overall prospective and achieve rapid overall reconstruction of the plane line position of the existing railway.
Owner:CENT SOUTH UNIV

Clustering method and system of mobile ad hoc network

The invention relates to a clustering method and system of a mobile ad hoc network. According to the clustering method and system, nodes are grouped, nodes in the same group satisfy a preset mobility condition, so that it can be ensured the nodes in the corresponding groups can have similar mobility, and therefore, the stability of clustering can be improved; the residual energy parameter values of the nodes, the node degrees of the nodes and the mobility parameter values of the nodes are calculated; the stability factors of the nodes are calculated; nodes corresponding to maximum stability factors in the groups are adopted as cluster heads; non-cluster head nodes in the groups are added into clusters where corresponding cluster heads are located; the nodes in the mobile ad hoc network are traversed, non-cluster head nodes which have not been clustered are clustered separately; and therefore, network node clustering can be realized. When the stability factors are calculated, the mobility parameter values of the nodes, the residual energy of the nodes and the node degrees of the nodes are all considered, and therefore, the cluster heads which are selected according to the stability factors of the nodes are optimal nodes with the mobility, residual energy and node degrees of the nodes considered; after the cluster heads are obtained, the non-cluster head nodes are clustered; and therefore, network clustering stability can be improved.
Owner:GUANGZHOU POWER SUPPLY BUREAU GUANGDONG POWER GRID CO LTD

Adaptive clustering method

InactiveCN105550244AAvoid the problem of difficult determination of weightImplementation of Cluster AdaptiveSpecial data processing applicationsText database clustering/classificationCluster algorithmSpatial cluster analysis
The invention discloses an adaptive clustering method. The method comprises the following five steps: processing spatial data and selecting characteristics; performing a spatial attribute clustering operation based on a Delaunay triangulation network; clustering by considering a non-spatial attribute; optimizing a clustering result; and visualizing the clustering result. The adaptive clustering method provided by the invention solves the problem of the traditional clustering algorithm that the clustering result is uncertain due to an uncertainty of parameter setting, has the characteristics of being capable of realizing automatic operation without prior knowledge, strong in adaptability, complete in function and so on, and can effectively improve the ability of analyzing and mining a deep geographical law by spatial clustering.
Owner:WUHAN UNIV

Resident customer clustering method and device based on demand response data

The invention discloses a resident customer clustering method and device based on demand response data, and the method comprises the steps: carrying out the power utilization census of resident users in a transformer area, and constructing a data matrix; according to the electricity utilization characteristics of the residential users, carrying out the dimension reduction processing on the electric quantity data of the users by using the related characteristic indexes; based on alpha-proximity and data encasement theories, providing a novel partitioning and layering clustering algorithm; analyzing the power consumption behaviors of the resident users before and after implementation based on a demand response incentive mechanism, and performing clustering analysis on the power consumption behaviors of the users. According to the device, nonvolatile software programs, instructions and modules in the memory are operated through the processor, so that various function applications and data processing of the server are executed, and clustering of resident customers is realized. According to the method, an aggregation theory method is applied to classification of resident users participating in demand response, and a scientific basis is provided for the process of customizing heterogeneous power plans for different users on a power grid side.
Owner:NANCHANG INST OF TECH

News and case correlation analysis method based on case element guidance and deep clustering

The invention relates to a news and case correlation analysis method based on case element guidance and deep clustering. The news and case correlation analysis method comprises the following steps: firstly, extracting important sentence representation texts; secondly, representing a case by using case elements to initialize a clustering center and guide a clustering search process; finally, selecting a convolutional auto-encoder to obtain text representation, using a reconstruction loss and clustering loss combined training network to enable the representation of the text to be closer to a case, unifying the text representation and clustering processes into the same frame, updating auto-encoder parameters and clustering model parameters alternately, and realizing text clustering. Aiming atthe problems that a current clustering algorithm lacks effective guidance information for news and case correlation analysis tasks, thus clustering divergence is caused, and the accuracy of a resultis reduced, the guidance effect of case elements in the clustering process and text vectorization representation is brought into full play, so the accuracy of a clustering result is effectively improved.
Owner:KUNMING UNIV OF SCI & TECH

Network security detection method and system

The invention discloses a network security detection method and system. The detection system comprises a central processing unit, a network virus detection module, a data acquisition module, a database module, a clustering module, a matching module, a message output module and a signal transmission module; the network virus detection module is connected with the central processing unit through thedata acquisition module; the central processing unit is respectively connected with the database module, the clustering module, the matching module and the message output module; and the central processing unit is connected with a background monitoring center through the signal transmission module. The network security detection method and system in the invention have a simple principle, can realize rapid detection of network viruses, can also achieve accurate classification of massive network data and improve the network security.
Owner:NANJING FOREST POLICE COLLEGE

File clustering method based on information bottleneck theory

The invention discloses a file clustering method based on an information bottleneck theory. The method firstly utilizes the information bottleneck theory to calculate the similarity between files; increment clustering algorithm is used for clustering files; minimum shared information loss is calculated on clustering result; if the minimum shared information loss satisfies a set threshold, the file is combined with the nearest cluster, otherwise a new cluster is created to store the file. Sequence clustering method is adopted for adjusting the clustering result to improve clustering accuracy, each file is sampled in sequence during adjusting process, and sampling frequency is set to control adjusting intensity. The adjusting policy contains all sample files and contributes to improving clustering accuracy.
Owner:BEIHANG UNIV

Streaming large-scale power data analysis method based on Spark Streaming

The invention discloses a streaming large-scale power data analysis method based on Spark Streaming, which comprises the following steps: 1, carrying out similarity search on online electric power data streams by utilizing an SS tree so as to cluster the electric power data; and 2, clustering the offline power data flow by using an improved Spark parallel K-means clustering method, wherein the clustering center of the K-means clustering and the initial value of the number of classes adopt the clustering center obtained in the step 1. Experimental evaluation of the method on a UCI data set shows that the method is superior to a traditional K-means clustering algorithm. Meanwhile, through testing the real data set of the user, it is found that electric power data of the user can be quickly and effectively clustered.
Owner:STATE GRID ZHEJIANG ELECTRIC POWER +1

Clustering method and device based on big data and electronic device

The invention provides a clustering method and device based on big data, and an electronic device. The method includes: obtaining a plurality of data points to be clustered and a preset number of clusters, in accordance with a first preset threshold, the distance between each of the data points and each of the other acquired data points, determining a density parameter for each of the data points,according to a density parameter of each of the data points, clustering a plurality of the data points into an initial cluster of the preset clustering number, and performing secondary clustering onthe plurality of the data points according to the plurality of the data points and each of the data points included in the initial cluster obtained by clustering to obtain a plurality of clusters of the preset clustering number. The invention can realize data clustering.
Owner:北信源系统集成有限公司

Virus sample clustering method and device, electronic equipment and storage medium

The invention discloses a virus sample clustering method, a virus sample clustering device, electronic equipment and a storage medium, which are used for clustering attacked viruses so as to process the attacked viruses and improve the safety. The clustering method comprises the following steps: acquiring behavior characteristics of a virus sample to be clustered; on the basis of a preset reference behavior feature set, generating a behavior identification group for the behavior features of the virus sample, wherein each reference behavior feature in the preset reference behavior feature set has a unique behavior identification; based on the behavior identification group of the virus sample, generating a group characteristic value for indicating a virus family to which the virus sample belongs; based on the group characteristic value of the virus sample, generating a family characteristic value used for indicating a virus family to which the virus sample belongs, each virus family comprising at least one virus family; and clustering the virus samples based on the group feature values and the ethnic group feature values of the virus samples.
Owner:TENCENT TECH (SHENZHEN) CO LTD

Adaptive pulse description word clustering method based on minimal value point of histogram

The invention relates to an adaptive clustering method for pulse description words at radar signal reconnaissance. The method comprises: two-dimensional frequency and pulse width clustering is carriedout on pulse description words; according to the frequency clustering and pulse width clustering methods, histogram statistics is carried out and a minimal value point in the histogram is searched instatistic results and is used as a boundary point of pulse description word groups, wherein the minimal value point of the histogram is defined as follows: if the height of a certain histogram rectangle is smaller than or equal to the height of a left histogram rectangle and is smaller than the height of a right histogram rectangle, the right boundary of the histogram rectangle is the minimal value point of the histogram.
Owner:THE 724TH RES INST OF CHINA SHIPBUILDING IND

Method and device for clustering sentences

The embodiment of the invention discloses a method and device for clustering sentences. One specific embodiment of the method comprises the steps of determining a set composed of semantic vectors corresponding to all sentences in a to-be-clustered sentence set as a semantic vector set; for each semantic vector in the semantic vector set, executing the following density calculation operation; for each semantic vector in the semantic vector set, executing the following clustering division operation; for each established cluster, determining the semantic vector with the maximum density in the semantic vectors divided into the cluster as the clustering center semantic vector of the cluster; and determining to-be-clustered sentences corresponding to the determined clustering center semantic vectors as a clustering center sentence set. According to the embodiment, the sentence clustering accuracy is improved.
Owner:BEIJING BAIDU NETCOM SCI & TECH CO LTD

Application list clustering method and device and computing equipment

The embodiment of the invention provides an application list clustering method and device and a piece of computing equipment. The method comprises the steps of obtaining application lists of a plurality of terminal equipment; for each application list, taking ordered application identifiers corresponding to the application lists as characteristic sets, and determining a fingerprint value corresponding to the characteristic set of each application list based on a Simhash algorithm; and classifying the application lists whose the fingerprint values meet clustering requirements as one category. According to the application list clustering method and device and the computing equipment provided by the invention, the purpose of clustering the application lists can be achieved with the relatively small data computing workload.
Owner:SHENZHEN TENCENT COMP SYST CO LTD

A Laplace centric peak data clustering method based on curvature

A Laplace centroid peak data clustering method based on curvature is disclosed. The data set to be classified is preprocessed to transform the data set to be classified into a weighted fully coupled network, and the Laplace centroid and the minimum distance value of the data points are calculated. The curvature-based method determines the optimal number of clusters R, and then chooses R data points with high Laplacian centrality and distance value as clustering centers. Finally, the remaining nodes are assigned and the clustering is completed. A method based on curvature determines that optimal clustering number, and can automatically find the correct number of clusters without predetermining the number of clusters, so as to realize the true parameterless clustering.
Owner:ZHEJIANG UNIV OF TECH

A method and apparatus for data cluster

A method and apparatus for data cluster are provided. The method comprises the steps of: in the case of obtaining uncertain data to be clustered, based on the uncertainty probability density functionof the uncertainty data, calculating the information needed for clustering the uncertain data, such as an based on uncertain probability density function of uncertain data, recalculating the preset initial centroid of the dataset, the sum of the expected square error of the recalculated preset initial centroid of the data set and the expected square error of the uncertain data to the preset initial centroid of the other data sets as uncertainty data considered as the sum of the expected square error of the uncertain data with respect to the data set, furthermore, determining the data set withthe minimum sum of expected square error as the target data set, and classifying the uncertain data into the target data set, so as to realize the clustering of the uncertain data by the uncertain probability density function based on the uncertain data, and improve the accuracy of the clustering of the uncertain data.
Owner:深圳软通动力科技有限公司

Clustering method, device and terminal

The invention discloses a clustering method, a clustering device and a clustering terminal. The method comprises the following steps of on the basis of a RCS (Rich Communication Suite) system, sending a clustering request containing clustering conditions to a server, and providing optional users which are fed back by the server and conform to the clustering conditions for an operating user; organizing a user group within the ranges of the operating user and the optional users which are selected by the operating user. A RCS terminal discloses by the invention interacts with the server on the basis of the RCS system, and the optional users which conform to the clustering conditions are obtained from a server end; one or more optional users can be selected by the user using the RCS terminal to form the user group together with the user to realize the clustering among the users, so that the experience effect of the user is improved.
Owner:COOLPAD SOFTWARE TECH (SHENZHEN) CO LTD

Data clustering method and device

The invention discloses a data clustering method and device. According to the invention, a local density clustering method is adopted, and a foundation is laid for clustering of sensor data through determining the distance between nodes, the local density of each node and the shortest distance between each node and a node with higher local density; and then the category to which each clustering center node and the nodes except for the clustering center nodes belong is determined according to the determined local density of each node and the shortest distance between each node and a node with higher local density, thereby realizing automatic clustering the nodes, accomplishing automatic extraction of data concepts, not only breaking through defects of the traditional k-means clustering method, but also realizing clustering for data of any shapes. In addition, the data clustering method and device lay a foundation for realizing collaborative analysis of heterogeneous equipment, interoperations of the equipment and the like, the reliability and complementarity of the information are ensured, and the accuracy of data concept extraction is improved.
Owner:CHINA MOBILE COMM LTD RES INST +1

A base station clustering method based on density and minimum distance in an ultra-dense network

The invention discloses a base station clustering method based on density and minimum distance in an ultra-dense network, comprising the following steps: firstly, calculating the distribution densityand clustering density threshold value of each micro-cell base station in the ultra-dense network, so that the micro-cell base stations of which the distribution density is greater than the clusteringdensity threshold value form an initial clustering center pool; calculating the minimum value of the distance between each micro-cell base station in the initial cluster center pool and the micro-cell base station with the distribution density higher than that of the micro-cell base station, defining the product of the distribution density of the micro-cell base stations and the minimum distanceas the weighted distribution density, and obtaining a to-be-selected cluster center pool according to the weighted distribution density; calculating a cluster center isolation distance, and sequentially removing the cluster center with a smaller weighted distribution density value in two cluster centers of which the distance between every two cluster centers is greater than the cluster center isolation distance in the to-be-selected cluster center pool from the to-be-selected cluster center pool; and finally, taking the number of cluster centers in the to-be-selected cluster center pool and the geographic position of the cluster center base station as parameters of traditional K-means algorithm, and executing K-means algorithm to obtain a clustering result. According to the method, the problem of non-uniform clustering is solved.
Owner:NANJING UNIV OF POSTS & TELECOMM

Time series data clustering method based on noise reduction encoder and attention mechanism

The invention provides a time series data clustering method based on a noise reduction encoder and an attention mechanism, which achieves clustering of label-free data in an unsupervised manner, and sequentially comprises the following steps of: 1, adding noise data into an original time sequence; step 2, adding an attention mechanism into the LSTM; 3, initializing a model by adopting a K-means method; 4, jointly optimizing the reconstruction loss and the KL divergence loss of the noise reduction auto-encoder by adopting an Adam optimization method; and iteratively optimizing the reconstruction loss and the clustering loss of the encoder to obtain a trained optimal model. Based on the deep noise reduction auto-encoder, the problem that time series data clustering is sensitive to noise is solved, and the clustering effect is improved by introducing an attention mechanism.
Owner:SOUTHEAST UNIV

Data clustering method and device, computer equipment and storage medium

The invention is suitable for the technical field of computers, and provides a data clustering method and device, computer equipment and a storage medium, and the method comprises the steps: respectively processing a plurality of pieces of clustering center data, and generating a plurality of corresponding core data sets; processing the remaining boundary data, respectively determining a core dataset associated with each piece of boundary data, and adding the core data sets into the corresponding core data sets; and performing clustering division on the to-be-clustered data according to the core data set. The data clustering method provided by the invention comprises the following steps of: firstly, dividing data which most possibly belongs to the same cluster into a core data set; compared with the prior art that the relevance between the remaining data and the divided data is determined, even if a certain point in the set is wrongly divided, the division of the remaining boundary data cannot be seriously influenced, and the technical problem that an existing clustering algorithm highly depends on the tag accuracy is solved.
Owner:重庆亿创西北工业技术研究院有限公司

Environment monitoring method and system based on Internet, storage medium and equipment

The invention relates to an environment monitoring method and system based on the Internet, a storage medium and equipment. The method comprises the steps of collecting environment parameter information of a target area in real time; making local space autocorrelation analysis on the target area when the environmental parameter information after signal processing triggers an early warning condition, and calculating a relative environmental quality index of a local space; and making minimum variance clustering analysis on the local spaces to obtain a relative environment quality grade of each type of local space. By preliminarily judging whether the early warning condition is triggered or not, the overall environment quality of the target area can be preliminarily evaluated, then local space autocorrelation analysis is conducted on the target area to conduct targeted analysis on the local space, clustering analysis is conducted in combination with the relative environment quality indexof the local space. The environmental quality grade of the local space of each category is determined, so that the environmental quality monitoring precision is greatly improved, the environmental quality condition of the target area can be reflected more accurately, and guidance is provided for scientific environmental governance.
Owner:兰州旭阳祥辉科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products