CN110276375B - Method for identifying and processing crowd dynamic clustering information - Google Patents
Method for identifying and processing crowd dynamic clustering information Download PDFInfo
- Publication number
- CN110276375B CN110276375B CN201910396365.3A CN201910396365A CN110276375B CN 110276375 B CN110276375 B CN 110276375B CN 201910396365 A CN201910396365 A CN 201910396365A CN 110276375 B CN110276375 B CN 110276375B
- Authority
- CN
- China
- Prior art keywords
- cluster
- population
- crowd
- algorithm
- clustering
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000012545 processing Methods 0.000 title claims abstract description 31
- 238000004458 analytical method Methods 0.000 claims abstract description 9
- 238000011156 evaluation Methods 0.000 claims description 21
- 238000011084 recovery Methods 0.000 claims description 11
- 238000005457 optimization Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000001953 sensory effect Effects 0.000 claims description 4
- 230000002068 genetic effect Effects 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 238000003875 gradient-accelerated spectroscopy Methods 0.000 claims description 2
- 238000002922 simulated annealing Methods 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 abstract 1
- 230000006870 function Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000002028 premature Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Probability & Statistics with Applications (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a method for identifying and processing crowd dynamic clustering information. The method solves the problem that the prior art is lack of identification and processing of the crowd dynamic clustering information. The method comprises the following steps: A. obtaining each cluster of the crowd dynamic information and solving the sum of the distance values from the individual in each cluster to the cluster center; B. calculating the sum of the distance values from the individual to the cluster center in each cluster again, and comparing the sum of the two distance values in the step A, B to judge whether the population fitness of the cluster is changed; C. calculating crowding distance values among individuals in each cluster to judge whether population diversity is deteriorated; and performing combined analysis on the results in the B and the C in a one-to-one correspondence manner, and performing targeted processing according to the analysis results. The method can effectively cluster the dynamic crowd information, analyze the crowd information according to the clustering condition, find the abnormal clustering condition in time and perform targeted adjustment processing.
Description
Technical Field
The invention belongs to the technical field of information processing, and relates to a method for identifying and processing crowd dynamic clustering information.
Background
Currently, various social software in China is continuously developed and widely applied, and brings great convenience to life and entertainment modes of people. However, the group-type illegal criminal behavior performed by using the social network has higher concealment and stronger destructiveness, which causes hidden troubles to the life and property safety of people. Therefore, by clustering the crowd data information in the social network, the personnel structures of different crowd groups and the relation of the crowd structures in the social big data network can be effectively mined.
Various types of network clustering algorithms have been introduced in the past decade, and the well-known algorithms include Girvan-Newman algorithm, rapid greedy module optimization, markov clustering algorithm, and the like. The detection of the crowd information in the network becomes an important research direction in the social data analysis, and the network data clustering method proposed by Orman and Forunto is applied to the crowd detection in the social network. However, because the network structure forms are different, complex and changeable, the traditional heuristic optimization method cannot satisfactorily solve the clustering and pattern recognition problem of the crowd information. Meta-heuristic algorithms have attracted a lot of attention from scholars in order to detect and identify crowd information. The algorithms have more remarkable advantages in terms of local learning and global searching capability, and show more remarkable optimization capability than the traditional heuristic algorithm in most researches. In the meta-heuristic algorithm, the crowd number can be automatically set by the algorithm, and online dynamic change can be realized. Many scholars have applied meta-heuristic algorithms to network clustering problems, such as evolutionary algorithms and particle swarm optimization algorithms. In the research, Pizzuti et al propose a single-target genetic algorithm for network clustering, and Gong proposes a Mernet algorithm based on the Memetic algorithm and a Memenet algorithm facing the network clustering problem. Although the above various single-target algorithms can cluster the crowd information, they cannot solve the recognition and processing problem of crowd information clustering in a dynamic environment well.
Disclosure of Invention
Aiming at the characteristics of the crowd information in the background technology, the invention provides a method for identifying and processing the crowd dynamic clustering information.
The purpose of the invention can be realized by the following technical scheme:
a method for identifying and processing crowd dynamic clustering information is characterized by comprising the following steps:
A. obtaining each cluster of the crowd dynamic information through multiple iterative computations, and solving the sum of the distance values from the individuals in each cluster to the cluster center;
B. and B, calculating the sum of the distance values from the individual to the cluster center in each cluster again, comparing the sum of the distance values with the sum of the distance values from the individual to the cluster center in the step A, judging that the population fitness of the cluster is changed when the sum of the two distance values is not equal, and judging that the population fitness of the cluster is not changed when the sum of the two distance values is equal:
C. calculating crowding distance values among individuals in each cluster to monitor the population diversity in the cluster and judge whether the population diversity is deteriorated or good;
D. and C, performing combined analysis processing on the results in the steps B and C in a one-to-one correspondence manner: when the combination result shows that the fitness of the cluster is unchanged and the population diversity is deteriorated, performing small-amplitude population diversity recovery operation on the cluster; when the combination result is that the fitness of the cluster is unchanged and the population diversity is good, no operation is performed on the cluster; when the combination result shows that the fitness of the cluster is changed and the population diversity is deteriorated, performing population diversity recovery operation on the cluster to a large extent; when the combination result indicates that the fitness of the cluster is changed and the population diversity is good, no operation is performed on the cluster;
in the above method for identifying and processing crowd dynamic clustering information, the clustering the crowd dynamic information includes the following steps:
a1, establishing a social graph model according to the social network of the crowd, wherein the model is represented as: g ═ G (N, V), where N is the number of nodes and V is the relationship between nodes;
a2, obtaining an adjacency matrix A according to the above model, the element a _ { ij } of the matrix being represented as:wherein L (i, j) represents that node i and node j are connected; w _ { ij } represents the weight of two nodes;
a3, obtaining the grade of the node iOrWherein the content of the first and second substances,wherein S is a clustering category of the crowd, and the social network is divided into m communities, then S ═ S1,S2,...,Sm},Is the in-degree of the ith node,is the out degree of the ith node;
a4, determining a strong sensory group and a weak sensory group by analyzing and judging the grade of the node i, thereby defining a clustering target of the crowd information, wherein the clustering target comprises a clustering target NRA and a clustering target RC:
a5 designing a meta-heuristic algorithm to cluster the clustering targets.
In the above method for identifying and processing crowd dynamic clustering information, the meta-heuristic algorithm includes the following steps:
a1, initializing population information;
a2, grouping the population information to form n sub-populations;
a3, performing one-to-one corresponding parallel operation on the n sub-populations by adopting n algorithms respectively;
a4, outputting an operation result, and evaluating the result of the parallel operation by using an evaluator;
a5, selecting an optimal algorithm according to the evaluation result, and carrying out iterative computation for a certain number of times by using the optimal algorithm;
a6, outputting the current overall sub-population, and judging whether each sub-population meets the termination condition of optimization: if yes, outputting a result; and if not, re-grouping the populations.
In the above method for identifying and processing dynamic crowd cluster information, the objective function of the crowd information cluster target includes: the hobbies, the places of the visitors, the habits and the social network structures of the crowds.
In the above method for identifying and processing crowd dynamic clustering information, in the parallel operation process, a recombiner of algorithms is designed to select a suitable algorithm, and the recombiner is used for recombining m suitable algorithms from n algorithms, where n is greater than m.
In the above method for identifying and processing crowd dynamic clustering information, the algorithm includes at least two of a particle swarm algorithm, an ant colony algorithm, a genetic algorithm, a simulated annealing algorithm, an artificial neural network algorithm and a GRASP with path-relining algorithm.
In the above method for identifying and processing crowd dynamic clustering information, the evaluator is configured to perform individual evaluation and sub-algorithm evaluation on the operation result; the individual evaluation is used for evaluating common individuals and better individuals in the operation result, and the sub-algorithm evaluation is used for performing algorithm evaluation on the better individuals.
In the above method for identifying and processing dynamic crowd clustering information, the crowding distance is calculated by using the following formula: the coordinates of two individuals in the population are respectively set as (x)1,x2,...,xn)、(y1,y2,...,yn) Then the distance between the two individuals is:
then the average distance of all individuals in the population is:
In the above method for identifying and processing dynamic crowd clustering information, the step C of determining the diversity of the crowd specifically includes: starting recording from the number of 1% of the allowable iteration number, assuming the iteration number is x, calculating the crowding distance of the population after the ith iteration is carried out (i is more than or equal to x), recording the crowding distance value di after the current iteration, and when the current iteration is carried outWhen the diversity is poor, the diversity is considered to be good.
In the above method for identifying and processing the dynamic crowd clustering information, in step D, when the analysis result indicates that the clustering needs to perform an operation of recovering the population diversity to a small extent or an operation of recovering the population diversity to a large extent, alarms of different levels are performed.
Compared with the prior art, the identification and processing method of the personal group dynamic clustering information has the following advantages: 1. the clustering of the crowd information is more accurate, and the dynamic crowd clustering information can be effectively monitored in time; 2. the crowd clustering management system can manage crowd clustering in time and perform key monitoring on the crowd structure with clustering.
Drawings
FIG. 1 is a schematic diagram of an analysis process according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of a social network of people in an embodiment of the present invention.
FIG. 3 is a framework for a meta-heuristic algorithm in an embodiment of the invention.
Detailed Description
The following are specific embodiments of the present invention and are further described with reference to the drawings, but the present invention is not limited to these embodiments.
As shown in fig. 1, the method for identifying and processing the personal group dynamic clustering information includes the following steps:
A. obtaining each cluster of the crowd dynamic information through multiple iterative computations, and solving the sum of the distance values from the individuals in each cluster to the cluster center;
B. and B, calculating the sum of the distance values from the individual to the cluster center in each cluster again, comparing the sum of the distance values with the sum of the distance values from the individual to the cluster center in the step A, judging that the population fitness of the cluster is changed when the sum of the two distance values is not equal, and judging that the population fitness of the cluster is not changed when the sum of the two distance values is equal:
C. calculating crowding distance values among individuals in each cluster to monitor the population diversity in the cluster and judge whether the population diversity is deteriorated or good;
D. and C, performing combined analysis processing on the results in the steps B and C in a one-to-one correspondence manner: when the combination result shows that the fitness of the cluster is unchanged and the population diversity is deteriorated, performing small-amplitude population diversity recovery operation on the cluster; when the combination result is that the fitness of the cluster is unchanged and the population diversity is good, no operation is performed on the cluster; when the combination result shows that the fitness of the cluster is changed and the population diversity is deteriorated, performing population diversity recovery operation on the cluster to a large extent; when the combination result indicates that the fitness of the cluster is changed and the population diversity is good, no operation is performed on the cluster;
specifically, as shown in fig. 2-3, clustering the crowd dynamic information includes the following steps: the crowd information clustering method based on the meta heuristic algorithm comprises the following steps:
a1, establishing a social graph model according to the social network of the crowd, wherein the model is represented as: g (N, V), wherein N is the number of nodes and corresponds to personnel information, and the dimensionality of N comprises characteristic information such as age, gender, ethnicity and the like; v is the relationship between nodes, which corresponds to the contact degree between different persons, and the information can be obtained by the relationship of persons registered in the case, and also can be obtained by analyzing and supplementing the big data of the person to and from.
A2, obtaining an adjacency matrix A according to the above model, the element a _ { ij } of the matrix being represented as:wherein L (i, j) represents that node i and node j are connected; w _ { ij } represents the weight of two nodes; in the clustering of the crowd information network, the characteristic similarity points of the nodes in the network need to be determined, and then information classification is carried out. If a network is indirect and unweighted when two nodes are connected, then a _ { ij } ═ 1; conversely, a _ { ij } -, 0.
A3, obtaining the grade of the node i:it can be specifically expressed as:wherein the content of the first and second substances,wherein S is a clustering category of the crowd, and the social network is divided into m communities, then S ═ S1,S2,...,Sm},Is the in-degree of the ith node,is the out degree of the ith node;
and A4, analyzing and judging the grade of the node i, and establishing a strong sense group and a weak sense group, wherein each node has more connections in the crowd than in other communities. In the weak population, the sum of the levels with the other population is greater than the sum of the levels within the population. Thereby defining a clustering target of the crowd information, wherein the clustering target comprises a clustering target NRA and a clustering target RC; NRA has a specific meaning of Negative Ratio Association, and RC has a specific meaning of Ratio Cut. The definition of the clustering objective is as follows:
a5, designing a meta-heuristic algorithm to perform clustering processing on the clustering target. The objective function of the crowd information clustering target comprises the following steps: the hobbies, the places of the visitors, the habits and the social network structures of the crowds. The meta-heuristic algorithm is an algorithm for scheduling and selecting various meta-heuristic algorithms by using a meta-heuristic algorithm idea, so that the adaptability of the algorithm to different problems is improved. As shown in fig. 2, the meta-heuristic algorithm employed in this step includes the following steps:
a51, initializing population information;
a52, grouping the population information to form n sub-populations;
a53, performing one-to-one corresponding parallel operation on the n sub-populations by adopting n algorithms respectively; and simultaneously, selecting a proper algorithm by using a recombiner, and recombining m proper algorithms from n algorithms by using the recombiner, wherein n is larger than m. Therefore, the modules of the meta-heuristic algorithm can be recombined and optimized, namely different algorithm operation modules are recombined as an algorithm module group (Modular Population/Swarm).
a54, outputting an operation result, and evaluating the result of the parallel operation by using an evaluator;
a55, selecting an optimal algorithm according to the evaluation result, and carrying out iterative computation for a certain number of times by using the optimal algorithm; the evaluator is used for carrying out individual evaluation and sub-algorithm evaluation on the operation result; the individual evaluation is used for evaluating common individuals and better individuals in the operation result, and the sub-algorithm evaluation is used for performing algorithm evaluation on the better individuals.
a56, outputting the current overall sub-population, and judging whether each sub-population meets the termination condition of optimization: if yes, outputting a result; and if not, re-grouping the populations.
The evaluator in this embodiment is an evaluation function, or objective function, of the candidate solution. For example, we want to minimize y — x1+ x2, and obtain two solutions, where the first solution is [ x1 — 1, x2 — 3], and the second solution is [ x1 — 2, x2 — 1 ].
Then in this problem, the evaluator is the objective function y x1+ x2, and for the first solution x1 1, x2 3, its evaluation result is 1+3 4, and for the second solution x1 2, x2 1, its evaluation result is 2+1 3)
The recombiner is to recombine the existing solution sets randomly, and according to the above example, the two solutions are respectively
Solution 1: [ x 1-1, x 2-3 ]
Solution 2: [ x 1-2, x 2-1 ]
Then a simpler and more convenient way to recombine is to interchange x1 in solution 1 with x1 in solution 2, generating two new solutions as follows:
new solution 1: [ x 1-1, x 2-1 ]
New solution 2: [ x1 ═ 2, x2 ═ 3]
And finally, evaluating the solutions 1 and 2, the new solutions 1 and 2 by using an evaluator, reserving the first two solutions of the quality, and entering the next iteration.
In step C, the congestion distance is calculated using the following formula: the coordinates of two individuals in the population are respectively set as (x)1,x2,...,xn)、(y1,y2,...,yn) Then the distance between the two individuals is:
then the average distance of all individuals in the population is:
The judgment of the population diversity specifically comprises the following steps: starting recording from the number of 1% of the allowable iteration number, assuming the iteration number is x, calculating the crowding distance of the population after the ith iteration is carried out (i is more than or equal to x), recording the crowding distance value di after the current iteration, and when the current iteration is carried outWhen the diversity is poor, the diversity is considered to be good.
And D, when the analysis result indicates that the cluster needs to be subjected to small-amplitude population diversity recovery operation or large-amplitude population diversity recovery operation, alarming at different levels. The operation of recovering population diversity tends to be good mainly by increasing the value of the distance between individuals, and the amplitude setting here is designed by human experience, but the average distance of all individuals can be referred to. Greater than 50% of the average distance may be considered significant. Distances less than 50% of the average distance may be considered small distances. .
The invention has the beneficial effects that:
1. and clustering by adopting the above-mentioned meta-heuristic algorithm. On one hand, the combination type of the sub-algorithms is further expanded in an exponential mode, so that the hyper-heuristic algorithm has stronger adaptivity; on the other hand, the design effectively reduces the spatial complexity of the algorithm (m sub-algorithms are selected from n algorithms to operate, but not all n sub-algorithms participate in the operation) while not increasing the time complexity of the algorithm (the maximum time complexity of each algorithm is not changed by the recombination of the algorithms). In the design of the cooperative algorithm structure, the original algorithm or learning mechanism in the high-level structure is abandoned, and the cooperative work of different algorithms is replaced.
2. The clustering abnormity of the dynamic crowd clustering information can be found in time: after the cluster information is evaluated again, if the fitness of the population is unchanged but the diversity of the population is deteriorated, the optimization environment is not changed, the deterioration of the diversity of the population is caused by algorithm convergence, and the recovery of the diversity of the population in a proper small amplitude is helpful for the algorithm to avoid premature and search stagnation; if the fitness of the population is unchanged and the diversity of the population is good, the algorithm does not operate on the diversity of the population, and the algorithm is continuously optimized in the current environment; if the fitness of the population is changed and the population diversity is worsened, the optimized environment is changed, and the population diversity is poorer, so the population diversity is greatly recovered; if the fitness of the population is changed, but the population diversity is good at the moment, the algorithm is enabled to continue to optimize in the current optimization environment, and the population diversity in the algorithm is not operated. Thus, the structure of the crowd with fusion can be monitored intensively, and potential dangers which may exist can be restrained early.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.
Claims (10)
1. A method for identifying and processing crowd dynamic clustering information is characterized by comprising the following steps:
A. obtaining each cluster of the crowd dynamic information through multiple iterative computations, and solving the sum of the distance values from the individuals in each cluster to the cluster center;
B. calculating the sum of the distance values from the individual to the cluster center in each cluster again, comparing the sum of the distance values with the sum of the distance values from the individual to the cluster center in the step A, judging that the population fitness of the cluster is changed when the sum of the two distance values is not equal, and judging that the population fitness of the cluster is not changed when the sum of the two distance values is equal;
C. calculating crowding distance values among individuals in each cluster to monitor the population diversity in the cluster and judge whether the population diversity is deteriorated or good;
D. and C, performing combined analysis processing on the results in the steps B and C in a one-to-one correspondence manner: when the combination result shows that the fitness of the cluster is unchanged and the population diversity is deteriorated, performing small-amplitude population diversity recovery operation on the cluster; when the combination result is that the fitness of the cluster is unchanged and the population diversity is good, no operation is performed on the cluster; when the combination result shows that the fitness of the cluster is changed and the population diversity is deteriorated, performing population diversity recovery operation on the cluster to a large extent; and when the combination result shows that the fitness of the cluster is changed and the population diversity is good, not performing any operation on the cluster.
2. The method as claimed in claim 1, wherein the step of clustering the dynamic crowd clustering information comprises the steps of:
a1, establishing a social graph model according to the social network of the crowd, wherein the model is represented as: g ═ G (N, V), where N is the number of nodes and V is the relationship between nodes;
a2, obtaining an adjacency matrix A according to the above model, the element a _ { ij } of the matrix being represented as:wherein L (i, j) represents that node i and node j are connected; w _ { ij } represents the weight of two nodes;
a3, obtaining the grade of the node iOrWherein the content of the first and second substances,wherein S is a clustering category of the crowd, and the social network is divided into m communities, then S ═ S1,S2,...,Sm},Is the in-degree of the ith node,is the out degree of the ith node;
a4, determining a strong sensory group and a weak sensory group by analyzing and judging the grade of the node i, thereby defining a clustering target of the crowd information, wherein the clustering target comprises a clustering target NRA and a clustering target RC:
a5 designing a meta-heuristic algorithm to cluster the clustering targets.
3. The method for identifying and processing crowd dynamic clustering information as claimed in claim 2, wherein the meta-heuristic algorithm comprises the steps of:
a51, initializing population information;
a52, grouping the population information to form n sub-populations;
a53, performing one-to-one corresponding parallel operation on the n sub-populations by adopting n algorithms respectively;
a54, outputting an operation result, and evaluating the result of the parallel operation by using an evaluator;
a55, selecting an optimal algorithm according to the evaluation result, and carrying out iterative computation for a certain number of times by using the optimal algorithm;
a56, outputting the current overall sub-population, and judging whether each sub-population meets the termination condition of optimization: if yes, outputting a result; and if not, re-grouping the populations.
4. The method for identifying and processing crowd dynamic clustering information according to claim 2 or 3, wherein the objective function of the crowd information clustering target comprises: the hobbies, the places of the visitors, the habits and the social network structures of the crowds.
5. The method as claimed in claim 3, wherein in the parallel operation process, a recombiner of algorithms is designed to select a suitable algorithm, and the recombiner is used to recombine m suitable algorithms from n algorithms, where n is greater than m.
6. The method as claimed in claim 3 or 5, wherein the algorithm includes at least two of a particle swarm algorithm, an ant colony algorithm, a genetic algorithm, a simulated annealing algorithm, an artificial neural network algorithm, and a GRASP with path-relining algorithm.
7. The method for identifying and processing crowd dynamic clustering information according to claim 3 or 5, wherein the evaluator is used for performing individual evaluation and sub-algorithm evaluation on the operation result; the individual evaluation is used for evaluating common individuals and better individuals in the operation result, and the sub-algorithm evaluation is used for performing algorithm evaluation on the better individuals.
8. A method for identifying and processing crowd dynamic clustering information according to claim 1, 2 or 3, wherein the crowd distance is calculated by the following formula: the coordinates of two individuals in the population are respectively set as (x)1,x2,...,xn)、(y1,y2,...,yn) Then the distance between the two individuals is:
then the average distance of all individuals in the population is:
9. The method for identifying and processing crowd dynamic clustering information according to claim 1, 2 or 3, wherein the judging of the crowd diversity in the step C specifically comprises: starting recording from the numerical number of 1% of the allowable iteration number, assuming that the iteration number is x, calculating the crowding distance of the population after the ith iteration is carried out (i is more than or equal to x), and recording the crowding distance value d after the current iterationiWhen is coming into contact withWhen the diversity is poor, the diversity is considered to be good.
10. The method for identifying and processing crowd dynamic clustering information according to claim 1, 2 or 3, wherein in step D, when the analysis result is that the clustering needs to perform a small-amplitude population diversity recovery operation or a large-amplitude population diversity recovery operation, different levels of alarm are performed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910396365.3A CN110276375B (en) | 2019-05-14 | 2019-05-14 | Method for identifying and processing crowd dynamic clustering information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910396365.3A CN110276375B (en) | 2019-05-14 | 2019-05-14 | Method for identifying and processing crowd dynamic clustering information |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110276375A CN110276375A (en) | 2019-09-24 |
CN110276375B true CN110276375B (en) | 2021-08-20 |
Family
ID=67959754
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910396365.3A Expired - Fee Related CN110276375B (en) | 2019-05-14 | 2019-05-14 | Method for identifying and processing crowd dynamic clustering information |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110276375B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117763228A (en) * | 2023-12-11 | 2024-03-26 | 广州小白信息技术有限公司 | Creative expression dynamic adaptation method based on multi-culture framework |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104537126A (en) * | 2015-01-29 | 2015-04-22 | 中南大学 | Overlapping community discovering method based on edge graph random walk |
CN108960486A (en) * | 2018-06-12 | 2018-12-07 | 郑州航空工业管理学院 | Interactive set evolvement method based on grey support vector regression prediction adaptive value |
CN109587144A (en) * | 2018-12-10 | 2019-04-05 | 广东电网有限责任公司 | Network security detection method, device and electronic equipment |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
IN2013MU02795A (en) * | 2013-08-27 | 2015-07-03 | Tata Consultancy Services Ltd | |
US9229095B2 (en) * | 2013-10-28 | 2016-01-05 | Raytheon Company | Adaptive dynamic cluster deinterleaving |
CN104731811B (en) * | 2013-12-20 | 2018-10-09 | 北京师范大学珠海分校 | A kind of clustering information evolution analysis method towards extensive dynamic short text |
CN108399267B (en) * | 2018-03-27 | 2020-04-14 | 东北大学 | Feedback clustering method based on cluster semantic feature analysis |
CN109685675A (en) * | 2019-01-11 | 2019-04-26 | 西南大学 | The didactic dynamic network community structure recognition methods of nature is propagated based on label |
-
2019
- 2019-05-14 CN CN201910396365.3A patent/CN110276375B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104537126A (en) * | 2015-01-29 | 2015-04-22 | 中南大学 | Overlapping community discovering method based on edge graph random walk |
CN108960486A (en) * | 2018-06-12 | 2018-12-07 | 郑州航空工业管理学院 | Interactive set evolvement method based on grey support vector regression prediction adaptive value |
CN109587144A (en) * | 2018-12-10 | 2019-04-05 | 广东电网有限责任公司 | Network security detection method, device and electronic equipment |
Non-Patent Citations (4)
Title |
---|
Design for a Novel Framework of Hyper-Heuristic Algorithm;GUO Wei-an,et al;《Journal of Donghua University》;20140831;第110页第5-9段 * |
Evolutionary Computation for Community Detection in Networks: A Review;Clara Pizzuti;《IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION》;20170809;第465页第3-5段、第472页第2段、第476页第7段 * |
基于多目标演化聚类的大规模动态网络社区检测;李赫等,;《计算机研究与发展》;20190228;第287页第5段、第288页第3段 * |
李赫等,.基于多目标演化聚类的大规模动态网络社区检测.《计算机研究与发展》.2019,281-292. * |
Also Published As
Publication number | Publication date |
---|---|
CN110276375A (en) | 2019-09-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wagh et al. | Survey on intrusion detection system using machine learning techniques | |
Cho | Incorporating soft computing techniques into a probabilistic intrusion detection system | |
CN111475804A (en) | Alarm prediction method and system | |
CN112085947A (en) | Traffic jam prediction method based on deep learning and fuzzy clustering | |
Louzada et al. | Bagging k-dependence probabilistic networks: An alternative powerful fraud detection tool | |
CN113269647B (en) | Graph-based transaction abnormity associated user detection method | |
Ramadan et al. | A novel hybrid intrusion detection system (IDS) for the detection of internet of things (IoT) network attacks | |
CN108009690A (en) | A kind of routine bus system optimized based on modularity steals group's automatic testing method | |
Liang et al. | ECF-MRS: An efficient and collaborative framework with Markov-based reputation scheme for IDSs in vehicular networks | |
Jain et al. | Singular race models: addressing bias and accuracy in predicting prisoner recidivism | |
CN110276375B (en) | Method for identifying and processing crowd dynamic clustering information | |
Abdullah et al. | An artificial deep neural network for the binary classification of network traffic | |
Ganapathy et al. | An intelligent system for intrusion detection using outlier detection | |
CN114897085A (en) | Clustering method based on closed subgraph link prediction and computer equipment | |
Pednekar et al. | Crime rate prediction using KNN | |
Khoshaim et al. | Emergency decision-making based on q-rung orthopair fuzzy rough aggregation information | |
CN113162919A (en) | Intrusion detection method based on network abnormal flow identification | |
Ageyev et al. | Traffic Abnormality Detection ML-based Method for IoT | |
CN110276376B (en) | Crowd information clustering method based on super element heuristic algorithm | |
Tran | Network anomaly detection | |
CN117134969A (en) | Intrusion detection algorithm based on diffusion generation countermeasure network and improved white whale optimization | |
CN116545679A (en) | Industrial situation security basic framework and network attack behavior feature analysis method | |
Rezaeipanah et al. | Intrusion detection in computer networks through combining particle swarm optimization and decision tree algorithms | |
Abadeh et al. | Using a particle swarm optimization approach for evolutionary fuzzy rule learning: a case study of intrusion detection | |
Narengbam et al. | Harris hawk optimization trained artificial neural network for anomaly based intrusion detection system |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210820 |