CN115587315A - Target self-adaptive identification method based on multi-group fusion - Google Patents
Target self-adaptive identification method based on multi-group fusion Download PDFInfo
- Publication number
- CN115587315A CN115587315A CN202211245453.1A CN202211245453A CN115587315A CN 115587315 A CN115587315 A CN 115587315A CN 202211245453 A CN202211245453 A CN 202211245453A CN 115587315 A CN115587315 A CN 115587315A
- Authority
- CN
- China
- Prior art keywords
- sample
- class
- clustering
- distance
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a target self-adaptive identification method based on multi-clustering fusion, which comprises the steps of firstly calculating local density rho and distance delta according to a density peak value clustering algorithm; then, a suitable clustering center is found according to a heuristic method of the decision graph, and a class label is distributed to each clustering center, namely, the clustering center is used as a class cluster and is marked as a positive class. Next, the remaining samples i in the data set are assigned to the class cluster where the sample j with higher local density and closest distance is located, and the samples i and j must be k neighbors. And if the samples i and j are not neighbors, assigning a new class label to the sample i, and marking as a negative class. The remaining samples may be assigned to the negative class in which sample i resides. When each sample assignment is completed, there may be some negative classes that need to be merged into the positive classes. The method can dynamically plan the clustering center and ensure the accuracy of target self-adaptive identification.
Description
Technical Field
The invention belongs to the field of target state identification, and particularly relates to a target self-adaptive identification method based on multi-group fusion.
Background
The problem of identifying the target state is to determine which state stored in the current knowledge base the currently input target radiation source signal sample corresponds to. For the problem, firstly, the samples of the current knowledge base are clustered and analyzed, and the class of the samples of the knowledge base is divided, wherein the application of a clustering algorithm is involved, and a density peak clustering algorithm (DPC) is a density-based clustering algorithm which can cluster clusters with any shape. On datasets with density differences between clusters of classes, DPC cannot accurately select the cluster center. The non-central point allocation strategy of the DPC can cause continuous errors and influence the clustering effect of the algorithm. The K-means algorithm is a classic clustering algorithm based on division, and takes distance as a standard for measuring similarity between data objects, that is, the smaller the distance between data objects is, the higher the similarity is, the more likely they are in the same class, but the disadvantage is that the number of clusters needs to be specified in advance, and the clustering result is sensitive to the selection of the initial class center.
Disclosure of Invention
In view of the above, the invention provides a target adaptive identification method based on multi-cluster fusion, which aims at the problem that a density peak clustering algorithm cannot accurately select a clustering center due to density differences among clusters, defines local density for eliminating the density differences among the clusters, establishes a distance judgment criterion, and realizes matching identification of the state of a target in a knowledge base.
In order to achieve the purpose, the technical scheme of the invention is realized as follows:
a target self-adaptive identification method based on multi-group fusion comprises the following steps,
performing clustering analysis on the samples of the current knowledge base by using a density peak value clustering algorithm, and dividing the classes of the samples of the knowledge base;
after the division result of each class is obtained, the distance between the current input sample and each clustering center is calculated, and a distance discrimination criterion is established to realize the matching identification of the state of the target in the knowledge base.
Further, the clustering analysis of the samples of the current knowledge base by using the density peak clustering algorithm to classify the classes of the samples of the knowledge base includes:
defining a neighbor set and a neighbor distance set of the samples, and calculating a distance matrix for all the samples in a knowledge base;
calculating the local density and the relative distance of the sample by using the distance matrix;
forming a decision graph by using the local density and the distance, determining clustering centers, and marking each clustering center as a positive class;
calculating a class cluster boundary;
allocating each sample except the cluster center to a cluster with high density and nearest sample, wherein the allocated sample must be the neighbor of the current input sample; otherwise, taking the current input sample as a newly generated clustering center, distributing a new class label, and marking as a negative class;
and calculating the cluster boundary of the classes, and merging the negative classes into the positive classes.
Further, after obtaining the partition results of each class, calculating the distance between the current input sample and each clustering center, establishing a distance discrimination criterion, and realizing the matching identification of the state of the target in the knowledge base, including:
inputting a current sample, calculating the distance from the current sample to each cluster center, and dividing the current sample into the categories which generate the minimum center distance.
Further, the method also comprises the following steps:
and updating the sample library, adding the current input into the sample library, and recalculating the clustering centers of various types.
Further, the method also comprises the following steps:
and judging whether the class containing the sample is removed, if the other clustering centers are not changed, finishing the target identification, wherein the identification result is the current class, and if the other clustering centers are changed, recalculating the distance from the current sample to each clustering center.
The invention also provides a target self-adaptive identification device based on the multi-grouping fusion, which comprises
The clustering analysis module is used for carrying out clustering analysis on the samples of the current knowledge base by using a density peak clustering algorithm and dividing the classes of the samples of the knowledge base;
and the target identification module is used for calculating the distance between the current input sample and each clustering center after the division results of each class are obtained, establishing a distance judgment criterion and realizing the matching identification of the state of the target in the knowledge base.
The present invention also provides an apparatus comprising a processor and a memory storing program instructions, the processor being configured to, upon execution of the program instructions, perform the above-described target adaptive identification method based on multi-cluster fusion.
The present invention also provides a computer readable storage medium, on which computer readable instructions are stored, the computer readable instructions being executed by a processor to implement the target adaptive identification method based on multi-group fusion.
Compared with the prior art, the target self-adaptive identification method based on the multi-class fusion has the following advantages that:
the invention realizes the matching identification of the state of the target in the knowledge base, effectively applies the clustering algorithm to the field of target identification, does not need iterative computation and updates the clustering center in the computation process, and effectively reduces the computation amount;
the invention provides a concept of updating a sample library, and simultaneously, the clustering result is analyzed, verified and corrected by using the change condition of the clustering center before and after the classification of the sample to be recognized, so that the accuracy of target self-adaptive recognition is ensured.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention, illustrate embodiments of the invention and together with the description serve to explain the invention and do not constitute a limitation of the invention. In the drawings:
FIG. 1 is a schematic flow chart of a target adaptive identification method based on multi-cluster fusion according to the present invention;
FIG. 2 is a schematic diagram of an adaptive target identification apparatus based on multi-cluster fusion according to the present invention;
fig. 3 is a schematic diagram of an electronic device of the present invention.
Detailed Description
It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict.
In the description of the present invention, it is to be understood that the terms "central," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like are used in the orientations and positional relationships indicated in the drawings, which are based on the orientations and positional relationships indicated in the drawings, and are used for convenience in describing the present invention and for simplicity in description, but do not indicate or imply that the device or element so referred to must have a particular orientation, be constructed in a particular orientation, and be operated, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or to implicitly indicate a number of the indicated technical features. Thus, a feature defined as "first," "second," etc. may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless otherwise specified.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood by those of ordinary skill in the art through specific situations.
The present invention will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
According to the analysis of the DPC algorithm and the k-means algorithm, a constraint condition is added to the fusion algorithm during sample distribution, the algorithm provided by the invention firstly refers to the DPC algorithm to calculate the local density rho and the distance delta; then, a suitable clustering center is found according to a heuristic method of the decision graph, and a class label is distributed to each clustering center, namely, the clustering center is used as a class cluster and is marked as a positive class. Then, the rest samples i in the data set are allocated to the class cluster where the sample j with higher local density and closest distance is located, and the samples i and j must be k neighbors. And if the samples i and j are not neighbors, assigning a new class label to the sample i, and marking as a negative class. The remaining samples may be assigned to the negative class in which sample i resides. When each sample allocation is complete, there may be some negative classes that need to be merged into the positive classes. The technical scheme is as shown in figure 1, and comprises the following steps:
step 1: defining a neighbor set N (i) of samples i and a neighbor distance set ND (i), calculating a distance matrix M for all samples in the knowledge base, wherein the neighbor set is defined as follows:
n (i) = { neighbor sample of j | i }
ND(i)＝{d(i,j)|j∈N(i)}
Step 2: calculating the local density and relative distance of the sample by using the distance matrix M
The formula is as follows:
ρ＝mean(D max -D est(i) )
wherein:
D max ＝max(N k D)
represents the maximum of all neighbor distances;
D est (i)＝{d(i,j)|d(i,j)<(D mean (i)+D median (i)/2)}
a neighbor set representing that the ith sample satisfies the constraint condition;
D mean (i)＝mean(ND(i))
means representing the nearest neighbor set of the ith sample;
D median (i)＝median(ND(i))
representing the median value of the ith sample neighbor set;
N k D＝{ND(1)∪ND(2)∪…,∪ND(n)}
representing a neighbor set of all samples.
And 3, step 3: forming a decision graph by using the local density and the distance, determining clustering centers, and recording each clustering center as a positive class;
and 4, step 4: calculating the cluster boundary, wherein the calculation formula is as follows:
edge(A P ,B N )＝{i|j∈N(i),j∈B N ,i∈A P }
wherein B is N 、A P Respectively representing the set of class clusters to which the samples i, j belong.
And 5: distributing each sample i except the clustering center to a cluster with the density higher than the local density of the cluster where the sample i is located and the closest sample j, wherein j must be the neighbor of i; otherwise, taking the sample i as a newly generated clustering center, distributing a new class label, and recording as a negative class;
step 6: repeating the calculation in the step 4, and combining the negative classes into the positive classes;
and 7: inputting a current sample, calculating the distance from the current sample to each clustering center, and dividing the distance into the categories generating the minimum center distance;
and 8: updating the sample library, adding the current input into the sample library, and recalculating the clustering centers of various types;
and step 9: if the class containing the sample is removed and other clustering centers are not changed, the target identification is finished, and the identification result is the current class; and if other cluster centers are changed, repeating the step 7-8.
The target self-adaptive identification algorithm of the multi-cluster fusion algorithm based on the fusion of the density peak value clustering algorithm and the k-means clustering algorithm can dynamically plan the clustering center and ensure the accuracy of target self-adaptive identification.
The invention also provides a target self-adaptive identification device based on the multi-cluster fusion, as shown in figure 2, the device comprises
The clustering analysis module is used for carrying out clustering analysis on the samples of the current knowledge base by using a density peak clustering algorithm and dividing the classes of the samples of the knowledge base;
and the target identification module is used for calculating the distance between the current input sample and each clustering center after the division result of each class is obtained, establishing a distance discrimination criterion and realizing the matching identification of the state of the target in the knowledge base.
The invention also provides an electronic device corresponding to the target adaptive identification method based on the multi-cluster fusion provided by the foregoing embodiment, so as to execute the method for target adaptive identification based on the multi-cluster fusion. As shown in fig. 3, which illustrates a schematic view of an electronic device provided by some embodiments of the present invention. The electronic device includes: the system comprises a processor, a memory, a bus and a communication interface, wherein the processor, the communication interface and the memory are connected through the bus; the memory stores a computer program that can be executed on the processor, and the processor executes the computer program to perform the method for adaptive object identification based on multi-group fusion provided by any of the foregoing embodiments of the present application. The Memory may include a Random Access Memory (RAM) and a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. The communication connection between the network element of the system and at least one other network element is realized through at least one communication interface (which may be wired or wireless), and the internet, a wide area network, a local network, a metropolitan area network and the like can be used. The bus may be an ISA bus, PCI bus, EISA bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory is used for storing a program, and the processor executes the program after receiving an execution instruction, and the target adaptive identification method based on multi-group fusion disclosed in any embodiment of the foregoing application may be applied to or implemented by the processor.
The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software. The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; but may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules may be located in ram, flash, rom, prom, or eprom, registers, etc. as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor.
The present invention further provides a computer-readable storage medium corresponding to the target adaptive identification method based on multi-cluster fusion provided in the foregoing embodiment, where the computer-readable storage medium is an optical disc, and a computer program (i.e., a program product) is stored on the optical disc, and when the computer program is executed by a processor, the computer program will execute the target adaptive identification method based on multi-cluster fusion provided in any of the foregoing embodiments. It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, a phase change memory (PRAM), a Static Random Access Memory (SRAM), a Dynamic Random Access Memory (DRAM), other types of Random Access Memories (RAM), a Read Only Memory (ROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a flash memory, or other optical and magnetic storage media, which are not described in detail herein.
All possible combinations of the technical features in the above embodiments may not be described for the sake of brevity, but should be considered as being within the scope of the present disclosure as long as there is no contradiction between the combinations of the technical features.
Claims (8)
1. A target self-adaptive identification method based on multi-cluster fusion is characterized in that:
performing clustering analysis on samples of the current knowledge base by using a density peak value clustering algorithm, and dividing the classes of the samples of the knowledge base;
after the division result of each class is obtained, the distance between the current input sample and each clustering center is calculated, and a distance discrimination criterion is established to realize the matching identification of the state of the target in the knowledge base.
2. The method of claim 1, wherein the target adaptive identification method based on the multi-class fusion is characterized in that: the method for clustering and analyzing the samples of the current knowledge base by using the density peak value clustering algorithm and dividing the classes of the samples of the knowledge base comprises the following steps:
defining a neighbor set and a neighbor distance set of the samples, and calculating a distance matrix for all the samples in a knowledge base;
calculating the local density and the relative distance of the sample by using the distance matrix;
forming a decision graph by using the local density and the distance, determining clustering centers, and recording each clustering center as a positive class;
calculating a class cluster boundary;
allocating each sample except the cluster center to a cluster with high density and nearest sample, wherein the allocated sample must be the neighbor of the current input sample; otherwise, taking the current input sample as a newly generated clustering center, distributing a new class label, and marking as a negative class;
and calculating the cluster boundary of the classes, and merging the negative classes into the positive classes.
3. The method of claim 1, wherein the target adaptive identification method based on the multi-class fusion is characterized in that: after the division results of all classes are obtained, the distance between the current input sample and each clustering center is calculated, a distance discrimination criterion is established, and the matching identification of the state of the target in the knowledge base is realized, and the method comprises the following steps:
inputting a current sample, calculating the distance from the current sample to each cluster center, and dividing the current sample into the categories which generate the minimum center distance.
4. The method of claim 1, wherein the target adaptive identification method based on the multi-class fusion is characterized in that: further comprising:
and updating the sample library, adding the current input into the sample library, and recalculating the clustering centers of various types.
5. The method of claim 1, wherein the target adaptive identification method based on the multi-class fusion is characterized in that: further comprising:
and judging whether the class containing the sample is removed, judging whether other clustering centers are changed or not, if not, finishing the target identification, wherein the identification result is the current class, and if the other clustering centers are changed, recalculating the distance from the current sample to each clustering center.
6. A target self-adaptive identification device based on multi-cluster fusion is characterized in that: comprises that
The clustering analysis module is used for carrying out clustering analysis on the samples of the current knowledge base by using a density peak clustering algorithm and dividing the classes of the samples of the knowledge base;
and the target identification module is used for calculating the distance between the current input sample and each clustering center after the division results of each class are obtained, establishing a distance judgment criterion and realizing the matching identification of the state of the target in the knowledge base.
7. An apparatus, characterized by: comprising a processor and a memory storing program instructions, the processor being configured, when executing the program instructions, to perform the method of any of claims 1 to 5.
8. A computer-readable storage medium: computer readable instructions stored thereon for execution by a processor to implement the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211245453.1A CN115587315A (en) | 2022-10-12 | 2022-10-12 | Target self-adaptive identification method based on multi-group fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211245453.1A CN115587315A (en) | 2022-10-12 | 2022-10-12 | Target self-adaptive identification method based on multi-group fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115587315A true CN115587315A (en) | 2023-01-10 |
Family
ID=84780469
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211245453.1A Pending CN115587315A (en) | 2022-10-12 | 2022-10-12 | Target self-adaptive identification method based on multi-group fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115587315A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116205236A (en) * | 2023-05-06 | 2023-06-02 | 四川三合力通科技发展集团有限公司 | Data rapid desensitization system and method based on entity naming identification |
-
2022
- 2022-10-12 CN CN202211245453.1A patent/CN115587315A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116205236A (en) * | 2023-05-06 | 2023-06-02 | 四川三合力通科技发展集团有限公司 | Data rapid desensitization system and method based on entity naming identification |
| CN116205236B (en) * | 2023-05-06 | 2023-08-18 | 四川三合力通科技发展集团有限公司 | Data rapid desensitization system and method based on entity naming identification |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107169768B (en) | Method and device for acquiring abnormal transaction data | |
| CN111476270B (en) | Course information determining method, device, equipment and storage medium based on K-means algorithm | |
| JP7238470B2 (en) | Learning device, inspection device, learning inspection method, learning program and inspection program | |
| CN106919957B (en) | Method and device for processing data | |
| CN115587315A (en) | Target self-adaptive identification method based on multi-group fusion | |
| CN108629375B (en) | Power customer classification method, system, terminal and computer readable storage medium | |
| CN111723815A (en) | Model training method, image processing method, device, computer system, and medium | |
| CN113191432B (en) | Outlier factor-based virtual machine cluster abnormality detection method, device and medium | |
| CN114417095A (en) | Data set partitioning method and device | |
| CN113837247A (en) | Product quality data clustering method | |
| CN115238788A (en) | Industrial big data classification method, system, server and storage medium | |
| CN111091022A (en) | Machine vision efficiency evaluation method and system | |
| CN111539576B (en) | Risk identification model optimization method and device | |
| CN111259806B (en) | Face area identification method, device and storage medium | |
| CN114091559A (en) | Data filling method and device, equipment and storage medium | |
| CN117152973B (en) | Expressway real-time flow monitoring method and system based on ETC portal data | |
| WO2022037198A1 (en) | Method and device for repairing fail bits | |
| CN115936736B (en) | Renewable resource recycling traceability evidence-preserving system and method | |
| WO2022198786A1 (en) | Target object detection method and apparatus, and electronic device and storage medium | |
| CN118069652A (en) | Method, device, equipment and storage medium for determining data stream information | |
| CN113469976A (en) | Object detection method and device and electronic equipment | |
| CN116610966A (en) | Data clustering method and device, processor and electronic equipment | |
| CN116343204A (en) | Flow cytometry cell classification method, system and terminal based on machine learning | |
| CN117391322A (en) | Method and device for matching capacity resources for waybills | |
| CN116776138A (en) | Interference data classification method, device and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |