CN117155786A - Directed network optimization method and system for screening robust influence nodes - Google Patents
Directed network optimization method and system for screening robust influence nodes Download PDFInfo
- Publication number
- CN117155786A CN117155786A CN202311003446.5A CN202311003446A CN117155786A CN 117155786 A CN117155786 A CN 117155786A CN 202311003446 A CN202311003446 A CN 202311003446A CN 117155786 A CN117155786 A CN 117155786A
- Authority
- CN
- China
- Prior art keywords
- population
- individuals
- node
- nodes
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000005457 optimization Methods 0.000 title claims abstract description 32
- 238000012216 screening Methods 0.000 title claims abstract description 30
- 238000011156 evaluation Methods 0.000 claims description 57
- 108090000623 proteins and genes Proteins 0.000 claims description 38
- 230000035772 mutation Effects 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 7
- 238000009825 accumulation Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 description 9
- 238000004590 computer program Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000002068 genetic effect Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000006467 substitution reaction Methods 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000007482 viral spreading Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0823—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Databases & Information Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a directed network optimization method and a directed network optimization system for screening robust influence nodes. The method is characterized in that according to the network structure of the directed network, the influence performance of each node when being attacked in the directed network is effectively evaluated, and a group of nodes with the maximum comprehensive influence performance is screened out.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a directed network optimization method and a directed network optimization system for screening robust influence nodes.
Background
Complex networks are widely found in people's daily lives and are widely used in modeling of social networks, traffic networks, the internet, etc. Members of the system can be modeled as a node with interconnections between nodes, the interconnections represented by links, the links having no direction, known as undirected networks, and having a direction, known as directed networks.
The process can be modeled as an impact maximization problem by selecting a set of nodes with the greatest impact from all nodes in a network. Taking the commodity popularization in the network as an example, the modes of advertising on the network social platform, such as WeChat, microblog, facebook and the like, become an efficient product popularization mode due to the convenience and universality of the network social platform. Due to the limitation of capital cost, it is impossible for a promoter to push promotion information to each user, while there are some more influential nodes in the social network, such as "UP main", "micro-doctor big V", etc., which can maximize the effect of product promotion given the cost. With the development of society, people are more and more closely connected, and the application value of the problem of maximizing influence is also more and more great.
Because of the complex environment often exposed, the network system is inevitably affected by interference and malicious attacks of external factors, and serious and interference and attacks can cause damage to the network structure and further affect the propagation of information on the network system. The impact of a node varies from node to node in the case of a damaged network to node, and the ability to maintain impact when the network is damaged is defined as the robust impact of the node. Under the constraint of a certain propagation model, a node with robust influence is selected to be modeled as a robust influence maximization problem. The information propagation performance depends on the integrity of a network structure and also depends on the anti-interference capability of a propagation node to external factors, and the problem of robust influence maximization is worth deeply discussing.
Disclosure of Invention
In view of the above, the embodiment of the invention provides a directed network optimization method and a directed network optimization system for screening robust influence nodes, which can efficiently perform directed network optimization for screening the robust influence nodes.
In one aspect, an embodiment of the present invention provides a directed network optimization method for screening robust influence nodes, including:
randomly selecting nodes of the target directed network to obtain a seed node set, and repeating the random node selection to obtain an initial population; wherein the initial population comprises a plurality of individuals, each individual representing a different set of seed nodes;
Taking the initial population as a first population, and performing cross operation on the first population to obtain a second population;
performing mutation operation on the second population to obtain a third population;
searching the third population, updating individuals in the third population, evaluating the fitness of the updated individuals, and obtaining a fourth population based on the evaluation result;
evaluating the fitness of all individuals of the fourth population, selecting individuals from the fourth population according to the evaluation result, performing population iteration, taking the iterated population as a first population, returning to the step of performing cross operation on the first population to obtain a second population until the preset iteration times are reached, and taking the population after the last iteration as a target population;
carrying out fitness evaluation on all individuals of the target population to obtain the individual with the largest robust influence in the target population as a target seed node set;
the first population, the second population, the third population, the fourth population and the target population all comprise a plurality of individuals, and each individual comprises a plurality of nodes in the target directional network; fitness assessment includes an assessment of the robust impact of the individual being assessed.
Optionally, the cross-over operation includes:
Exchanging nodes of two random individuals in the first population based on preset gene points;
after the crossover operation, when a repeated node exists in the individual of the exchanged nodes, a new node is obtained from the target directed network to replace the repeated node.
Optionally, the mutation operation includes:
obtaining a new node from the target directed network to replace a random node of any individual in the second population;
wherein the new node does not repeat with any node in the individual of the replaced node.
Optionally, performing a search operation on the third population, updating individuals in the third population, and further performing fitness evaluation on the updated individuals, and obtaining a fourth population based on the evaluation result, including:
performing global search operation on the first individuals in the third population, performing node replacement on the first individuals in the third population through target nodes in the target directed network, further performing fitness evaluation on the first individuals before and after the node replacement respectively, and when the robustness influence of the first individuals after the node replacement is improved, reserving the first individuals after the node replacement;
and/or performing neighborhood searching operation on the second individuals in the third population, performing node replacement on the second individuals through traversing and searching neighborhood nodes of all nodes in the second individuals in the third population, performing fitness evaluation on the second individuals before and after the node replacement by using random neighborhood nodes, and when the robust influence of the second individuals after the node replacement is improved, retaining the second individuals after the node replacement;
And/or performing random search operation on a third individual in the third population, performing node replacement on the third individual by randomly acquiring new nodes from the target directed network, further performing fitness evaluation on the third individual before and after the node replacement, and when the robust influence of the third individual after the node replacement is improved, reserving the third individual after the node replacement;
and obtaining a fourth population according to the individuals of the third population remained after node replacement.
Optionally, performing fitness evaluation on all individuals of the fourth population, and further selecting individuals from the fourth population according to the evaluation result to perform population iteration, including:
carrying out fitness evaluation on all individuals of the fourth population to obtain the robust influence of all the individuals of the fourth population;
sorting all individuals of the fourth population according to the level of the robust influence, screening the individuals of the new population of the sorted fourth population through sequential sorting selection and roulette selection, and further carrying out population iteration according to the screened individuals of the new population.
Optionally, the fitness evaluation includes:
continuously simulating network attack on the target directed network, removing the attacked nodes in the target directed network one by one, and further determining the robust influence of the evaluated individual according to the influence level of the evaluated individual in the network structure of the removed nodes of the target directed network in each stage and the control link number of the network structure of the target directed network in each stage of the removed nodes.
Optionally, determining the robust influence of the evaluated individual according to the influence level of the evaluated individual in the network structure of the target directed network at each stage of the removal node and the control link number of the network structure of the target directed network at each stage of the removal node comprises:
determining an influence level from the difference between the sum of the influence ranges of the evaluated individuals in the network structure of each stage of the removed node and the overlapping influence;
determining state values of the evaluated individuals at each stage of removing nodes of the target directional network according to the ratio of the influence level to the control link number;
and carrying out accumulation normalization processing on the state values of the evaluated individuals in all stages of the removal node of the target directional network to obtain the robust influence of the evaluated individuals on the target directional network.
In another aspect, an embodiment of the present invention provides a directed network optimization system for screening robust influence nodes, including:
the first module is used for carrying out node random selection on the target directed network to obtain a seed node set, and repeating the node random selection to obtain an initial population; wherein the initial population comprises a plurality of individuals, each individual representing a different set of seed nodes;
The second module is used for taking the initial population as a first population, performing cross operation on the first population and obtaining a second population;
the third module is used for carrying out mutation operation on the second population to obtain a third population;
the fourth module is used for carrying out search operation on the third population, updating individuals in the third population, further carrying out fitness evaluation on the updated individuals, and obtaining a fourth population based on an evaluation result;
a fifth module, configured to perform fitness evaluation on all individuals of the fourth population, further select individuals from the fourth population according to the evaluation result, perform population iteration, use the iterated population as a first population, and then return to the second module to perform cross operation on the first population, so as to obtain a second population, until reaching a preset iteration number, and use the population after the last iteration as a target population;
a sixth module, configured to perform fitness evaluation on all individuals in the target population, and obtain an individual with the largest robust influence in the target population as a target seed node set;
the first population, the second population, the third population, the fourth population and the target population all comprise a plurality of individuals, and each individual comprises a plurality of nodes in the target directional network; fitness assessment includes an assessment of the robust impact of the individual being assessed.
In another aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory;
the memory is used for storing programs;
the processor executes a program to implement the method as before.
In another aspect, embodiments of the present invention provide a computer-readable storage medium storing a program for execution by a processor to perform a method as previously described.
Embodiments of the present invention also disclose a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The computer instructions may be read from a computer-readable storage medium by a processor of a computer device, and executed by the processor, to cause the computer device to perform the foregoing method.
The method comprises the steps of firstly, randomly selecting nodes of a target directional network to obtain a seed node set, and repeating the random node selection to obtain an initial population; wherein the initial population comprises a plurality of individuals, each individual representing a different set of seed nodes; taking the initial population as a first population, and performing cross operation on the first population to obtain a second population; performing mutation operation on the second population to obtain a third population; searching the third population, updating individuals in the third population, evaluating the fitness of the updated individuals, and obtaining a fourth population based on the evaluation result; evaluating the fitness of all individuals of the fourth population, selecting individuals from the fourth population according to the evaluation result, performing population iteration, taking the iterated population as a first population, returning to the step of performing cross operation on the first population to obtain a second population until the preset iteration times are reached, and taking the population after the last iteration as a target population; carrying out fitness evaluation on all individuals of the target population to obtain the individual with the largest robust influence in the target population as a target seed node set; the first population, the second population, the third population, the fourth population and the target population all comprise a plurality of individuals, and each individual comprises a plurality of nodes in the target directional network; fitness assessment includes an assessment of the robust impact of the individual being assessed. According to the embodiment of the invention, through the fitness evaluation of the robust influence and the combination of the loop iteration optimization, the directed network optimization for screening the robust influence nodes can be efficiently performed.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of a directed network optimization method for screening robust influence nodes according to an embodiment of the present invention;
fig. 2 is an overall flow diagram of a directed network optimization method for screening robust influence nodes according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a directed network optimization system for screening robust influence nodes according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a frame of an electronic device according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
In one aspect, as shown in fig. 1, an embodiment of the present invention provides a directed network optimization method for screening robust influence nodes, including:
s100, randomly selecting nodes of a target directional network to obtain a seed node set, and repeating the random selection of the nodes to obtain an initial population;
wherein the initial population comprises a plurality of individuals, each individual representing a different set of seed nodes;
in some embodiments, an initialization operation is performed to randomly select K nodes from the network as a seed node set and become a single individual. The random selection operation is repeated until a population is formed having a certain number of individuals. Generating an initial population P by executing an initialization operator 0 And the subsequent genetic evolution operation is convenient.
S200, taking the initial population as a first population, and performing cross operation on the first population to obtain a second population;
in some embodiments, the cross operation may include: exchanging nodes of two random individuals in the first population based on preset gene points; after the crossover operation, when a repeated node exists in the individual of the exchanged nodes, a new node is obtained from the target directed network to replace the repeated node.
In some embodiments, crossover operations are performed, population P t Every two individuals in the population cross, randomly extracting two individuals for cross until all individuals finish cross, and combining the sub-individuals generated after cross and the original parent individuals into a new population cross: and selecting a gene point, exchanging the front end and the rear end of the genes of the two individuals, and randomly selecting a replacement from the nodes in the network if repeated nodes are generated in the crossed genes, so as to ensure that no repeated nodes exist in the genes.
S300, performing mutation operation on the second population to obtain a third population;
in some embodiments, the mutation operation may include: obtaining a new node from the target directed network to replace a random node of any individual in the second population; wherein the new node does not repeat with any node in the individual of the replaced node.
In some embodiments, the mutation is performed on the crossed population P t ′ Each individual randomly selects a node from the network to replace a node in the gene given the probability of variation. The mutation operation aims at maintaining the genetic diversity of the population and enriching the number of the node sets of the seed to be selected.
S400, searching the third population, updating individuals in the third population, evaluating fitness of the updated individuals, and obtaining a fourth population based on an evaluation result;
it should be noted that, in some embodiments, step S400 may include: performing global search operation on the first individuals in the third population, performing node replacement on the first individuals in the third population through target nodes in the target directed network, further performing fitness evaluation on the first individuals before and after the node replacement respectively, and when the robustness influence of the first individuals after the node replacement is improved, reserving the first individuals after the node replacement; and/or performing neighborhood searching operation on the second individuals in the third population, performing node replacement on the second individuals through traversing and searching neighborhood nodes of all nodes in the second individuals in the third population, performing fitness evaluation on the second individuals before and after the node replacement by using random neighborhood nodes, and when the robust influence of the second individuals after the node replacement is improved, retaining the second individuals after the node replacement; and/or performing random search operation on a third individual in the third population, performing node replacement on the third individual by randomly acquiring new nodes from the target directed network, further performing fitness evaluation on the third individual before and after the node replacement, and when the robust influence of the third individual after the node replacement is improved, reserving the third individual after the node replacement; and obtaining a fourth population according to the individuals of the third population remained after node replacement.
It should be noted that the global search operation, the neighborhood search operation, and the random search operation described above are performed with a certain probability in each individual in the third population, and the first individual represents a plurality of individuals in the third population that perform the global search operation (theoretically, each individual may perform the global search operation), and similarly, the second individual represents a plurality of individuals in the third population that perform the neighborhood search operation (theoretically, each individual may perform the neighborhood search operation), and the third individual represents a plurality of individuals in the third population that perform the random search operation (theoretically, each individual may perform the random search operation).
In some embodiments, the search operations include a global search operation, a neighborhood search operation, and a random search operation, each of the search operations having the steps of:
and (3) global searching operation, wherein a global searching operator is executed with a certain probability for each individual in the population obtained in the previous processing step. And selecting a certain node with the top 2% of the degree ranking in the global network, replacing a certain gene in an individual, and recalculating the fitness of the gene. If the fitness of the individual after replacement is improved, the replaced genes are reserved, otherwise, the genes are not reserved.
And (3) carrying out a neighborhood searching operation, wherein a neighborhood searching operator is executed with a certain probability for each individual in the population obtained in the previous processing step. The selected individual will search through all nodes in the neighborhood with each node distance of 2, randomly select one of the neighborhood nodes for replacement, and recalculate its fitness. If the fitness of the individual after replacement is improved, the replaced genes are reserved, otherwise, the genes are not reserved.
And (3) a random search operation, wherein a random search operator is executed with a certain probability for each individual in the population obtained in the previous processing step. A node is randomly selected from the network to replace a node in the gene and its fitness is recalculated. If the fitness of the individual after replacement is improved, the replaced genes are reserved, otherwise, the genes are not reserved.
S500, carrying out fitness evaluation on all individuals of a fourth population, selecting individuals from the fourth population according to an evaluation result to carry out population iteration, taking the iterated population as a first population, returning to the step of carrying out cross operation on the first population to obtain a second population until the preset iteration times are reached, and taking the population after the last iteration as a target population;
it should be noted that in some embodiments, performing fitness evaluation on all the individuals of the fourth population, and further selecting the individuals from the fourth population to perform population iteration according to the evaluation result may include: carrying out fitness evaluation on all individuals of the fourth population to obtain the robust influence of all the individuals of the fourth population; sorting all individuals of the fourth population according to the level of the robust influence, screening the individuals of the new population of the sorted fourth population through sequential sorting selection and roulette selection, and further carrying out population iteration according to the screened individuals of the new population.
In some embodiments, the selection operation is performed first, and the screening of the individuals of the new population is performed according to a selection method of mixing sequential ranking selection and roulette selection according to the fitness. The fitness (robustness) size is ranked from high to low for the old population. The first half of the names in the new population are directly selected by the old population according to the ranking, the other half of the names are selected from the unselected individuals in the old population by means of roulette, the probability of being selected into the new population is higher the more the ranking is, and the individuals with smaller fitness also have the opportunity of being selected. The selection operation ensures the optimal gene retention and improves the diversity of the genes of the new population. And further, the population iteration is circularly carried out, and the related operations of the steps S200 to S500 are repeatedly executed until the set termination condition is met.
S600, carrying out fitness evaluation on all individuals of the target population to obtain the individual with the largest robust influence in the target population as a target seed node set;
it should be noted that, in some embodiments, the fitness evaluation may include: continuously simulating network attack on the target directed network, removing the attacked nodes in the target directed network one by one, and further determining the robust influence of the evaluated individual according to the influence level of the evaluated individual in the network structure of the removed nodes of the target directed network in each stage and the control link number of the network structure of the target directed network in each stage of the removed nodes.
Wherein in some embodiments, determining the robust impact of the evaluated individual according to the impact level of the evaluated individual in the network structure of the target directed network at the various stages of the removal node and the control link number of the network structure of the target directed network at the various stages of the removal node may comprise: determining an influence level from the difference between the sum of the influence ranges of the evaluated individuals in the network structure of each stage of the removed node and the overlapping influence; determining state values of the evaluated individuals at each stage of removing nodes of the target directional network according to the ratio of the influence level to the control link number; and carrying out accumulation normalization processing on the state values of the evaluated individuals in all stages of the removal node of the target directional network to obtain the robust influence of the evaluated individuals on the target directional network.
In some embodiments, the directed network node robust impact performance assessment index is as follows:
n represents the number of all nodes in a network; q represents the proportion of nodes attacked by the outside to all nodes; sigma (sigma) S (q) represents removal of qNAfter nodes, a group of nodes S have influence levels in the network structure formed by the rest nodes, and the calculation method comprises the following steps of Calculating, wherein a first term represents the sum of influence ranges of which the distance of each node in the seed node set S is 1; the second term minus the overlapping influence of the nodes of the seed node set S over a distance of 1, set C s Is the neighboring node within the distance 1 of the seed s, p (c, s) is the propagation probability between the active node s and the node c to be activated, +.>Is the direct neighbor of node c; the third term χ minus is the overlapping influence of the newly activated node c on the seed set S, specifically +.>N D (q) represents the number of control links in the directed network structure consisting of the remaining nodes after the q×n nodes are shifted out.
The flow of the node robust influence evaluation when the directed network is attacked by the outside is as follows:
(1) The attack process simulates an attack on a node, and for the node under attack, the operation of moving out of the node is to move it out of the network and disconnect the links of the node with all other neighboring nodes. For the order of removing the nodes, specifically: and sequencing all nodes in the network from high to low according to degrees (assuming that an attacker preferentially attacks the node with the greatest degree in the network), and sequentially selecting the node with the greatest degree to perform the shifting-out operation.
(2) Every time a node is moved out, according to the network structure formed by the remaining nodes after the node is moved out and the candidate node set, calculating an independent state according to a formula After all nodes are moved out, R in all states is carried out S Performing accumulation normalization processing to obtain robust influence index +.>
In some embodiments, as shown in fig. 2, the method performs optimal seed node set screening based on a modulo-cause algorithm, and the flow is as follows:
(a) Executing an initialization operator to generate a population with a certain size, wherein each population comprises a plurality of individuals, each individual is a candidate node set S, and the nodes of each node set are randomly extracted from a network;
(b) Evaluating the fitness of all individuals in the population, and evaluating the fitness of each individual by taking the robust influence as an index;
(c) Executing a crossover operator, exchanging individual genes in the population, generating new genes and the population, and expanding the number of candidate seed nodes;
(d) Executing a mutation operator to randomly generate new genes for individuals in the population, so as to enrich the genetic diversity of the population;
(e) Executing a global searching operator, searching nodes with larger degrees in the global scope of the network, and taking the more optimal nodes as genes in the gene optimization population;
(f) Executing a neighborhood searching operator, traversing partial nodes in the neighborhood of the candidate node, and selecting a better node as a gene of the original individual of gene optimization;
(g) Executing a random search operator, randomly selecting nodes as genes to replace original genes in the global range of the network, and reserving if higher adaptability is generated;
(h) Executing a selection operator, and screening all individuals in the population for proper individuals according to the fitness to enter the next generation;
(i) Repeating the operations (b) - (h) until a set number of iterations is met and an optimal individual robust impact level is output.
There are many studies and methods for the problem of maximizing impact and the problem of maximizing robust impact. But these studies have focused mainly on undirected networks, lacking in the directionality of node propagation and in-depth study of the characteristics of the directed networks. In real life, the propagation direction such as the road traffic direction or the virus spreading direction has important significance for the propagation model. In addition, the existing node robust influence optimization methods such as heuristic methods, monte Carlo process methods and the like have the defects of low convergence speed and low optimization accuracy. In view of the shortcomings of the existing methods, it is necessary to design an index for effectively evaluating the influence performance of nodes in a directed network, and to provide a seed node screening method with both robustness and influence to solve the problem of maximizing the influence in the directed network.
The invention discloses an index for measuring the robustness and the comprehensive performance of influence of nodes in a directed network under the condition that the directed network is attacked by the outside, solves the evaluation problem of the stable information transmission capability of the nodes in the directed network, and designs a method for solving the problem of the maximization of the robustness influence in the directed network by a modulo factor algorithm based on the index. Robust impact performance assessment index for directed network nodes:
wherein N represents the number of all nodes in a network, q represents the proportion of nodes attacked by the outside to all nodes, and sigma S (q) represents the level of influence of a group of nodes S in the network structure of the remaining nodes after the q N nodes are removed, N D (q) represents the number of control links in the directed network structure consisting of the remaining nodes after the q×n nodes are shifted out.
The optimal seed node set screening method based on the modulo-cause algorithm is as follows: executing an initialization operator to generate a population with a certain size, and generating an initial population, so that the subsequent genetic evolution operation is facilitated; executing a crossover operator, exchanging individual genes in the population, generating new genes and the population, and expanding the number of candidate seed nodes; executing a mutation operator, randomly generating new genes, and enriching the diversity of population genes; executing a global searching operator, searching nodes with larger degrees in the global scope of the network, and taking the more optimal nodes as genes in the gene optimization population; executing a neighborhood searching operator, traversing nodes in the neighborhood of the candidate node, and selecting a better node as a gene of the original individual of gene optimization; executing a random search operator, and randomly selecting a better node from a global range as a gene to replace an original gene; executing a selection operator, and screening all individuals in the population for proper individuals according to the fitness to enter the next generation; the above operations are repeatedly performed until the set number of iterations is satisfied and the most individual robust impact level is output.
In another aspect, as shown in fig. 3, an embodiment of the present invention provides a directed network optimization system 700 for screening robust influence nodes, including: a first module 710, configured to perform node random selection on a target directional network to obtain a seed node set, and repeat node random selection to obtain an initial population; wherein the initial population comprises a plurality of individuals, each individual representing a different set of seed nodes; a second module 720, configured to take the initial population as a first population, perform a cross operation on the first population, and obtain a second population; a third module 730, configured to perform a mutation operation on the second population to obtain a third population; a fourth module 740, configured to perform a search operation on the third population, update individuals in the third population, further perform fitness evaluation on the updated individuals, and obtain a fourth population based on the evaluation result; a fifth module 750, configured to perform fitness evaluation on all individuals of the fourth population, further select individuals from the fourth population according to the evaluation result to perform population iteration, take the iterated population as a first population, and then return to the second module to perform cross operation on the first population to obtain a second population, until reaching a preset iteration number, and take the population after the last iteration as a target population; a sixth module 760, configured to perform fitness evaluation on all the individuals in the target population, and obtain the individual with the largest robust influence in the target population as the target seed node set; the first population, the second population, the third population, the fourth population and the target population all comprise a plurality of individuals, and each individual comprises a plurality of nodes in the target directional network; fitness assessment includes an assessment of the robust impact of the individual being assessed.
The content of the method embodiment of the invention is suitable for the system embodiment, the specific function of the system embodiment is the same as that of the method embodiment, and the achieved beneficial effects are the same as those of the method.
As shown in fig. 4, another aspect of an embodiment of the present invention also provides an electronic device 800, including a processor 810 and a memory 820;
the memory 820 is used for storing programs;
processor 810 executes a program to implement the method as before.
The content of the method embodiment of the invention is suitable for the electronic equipment embodiment, the functions of the electronic equipment embodiment are the same as those of the method embodiment, and the achieved beneficial effects are the same as those of the method.
Another aspect of the embodiments of the present invention also provides a computer-readable storage medium storing a program that is executed by a processor to implement a method as before.
The content of the method embodiment of the invention is applicable to the computer readable storage medium embodiment, the functions of the computer readable storage medium embodiment are the same as those of the method embodiment, and the achieved beneficial effects are the same as those of the method.
Embodiments of the present invention also disclose a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The computer instructions may be read from a computer-readable storage medium by a processor of a computer device, and executed by the processor, to cause the computer device to perform the foregoing method.
In some alternative embodiments, the functions/acts noted in the block diagrams may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may in fact be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Furthermore, the embodiments presented and described in the flowcharts of the present invention are provided by way of example in order to provide a more thorough understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed, and in which sub-operations described as part of a larger operation are performed independently.
Furthermore, while the present invention has been described in the context of functional modules, it should be appreciated that, unless otherwise indicated, one or more of the functions and/or features of the present invention may be integrated in a single physical device and/or software module or may be implemented in separate physical devices or software modules. It will also be appreciated that a detailed discussion of the actual implementation of each module is not necessary to an understanding of the present invention. Rather, the actual implementation of the various functional modules in the apparatus disclosed herein will be apparent to those skilled in the art from consideration of their attributes, functions and internal relationships. Accordingly, one of ordinary skill in the art can implement the invention as set forth in the claims without undue experimentation. It is also to be understood that the specific concepts disclosed are merely illustrative and are not intended to be limiting upon the scope of the invention, which is to be defined in the appended claims and their full scope of equivalents.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution apparatus, device, or apparatus, such as a computer-based apparatus, processor-containing apparatus, or other apparatus that can fetch the instructions from the instruction execution apparatus, device, or apparatus and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution apparatus, device, or apparatus.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium upon which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution device. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that: many changes, modifications, substitutions and variations may be made to the embodiments without departing from the spirit and principles of the invention, the scope of which is defined by the claims and their equivalents.
While the preferred embodiment of the present invention has been described in detail, the present invention is not limited to the embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention, and the equivalent modifications or substitutions are intended to be included in the scope of the present invention as defined in the appended claims.
Claims (10)
1. A directed network optimization method for screening robust influence nodes, comprising:
randomly selecting nodes of a target directed network to obtain a seed node set, and repeating the random node selection to obtain an initial population; wherein said initial population comprises a plurality of individuals, each of said individuals characterizing a different set of said seed nodes;
taking the initial population as a first population, and performing cross operation on the first population to obtain a second population;
performing mutation operation on the second population to obtain a third population;
Searching the third population, updating individuals in the third population, evaluating the fitness of the updated individuals, and obtaining a fourth population based on an evaluation result;
performing fitness evaluation on all individuals of the fourth population, selecting individuals from the fourth population according to an evaluation result to perform population iteration, taking the iterated population as a first population, returning to the step of performing cross operation on the first population to obtain a second population until the preset iteration times are reached, and taking the population after the last iteration as a target population;
carrying out the fitness evaluation on all individuals of the target population to obtain the individual with the largest robust influence in the target population as a target seed node set;
wherein the first population, the second population, the third population, the fourth population, and the target population each comprise a plurality of individuals, each individual comprising a number of nodes in the target-directed network; the fitness assessment includes an assessment of the robust impact of the assessed individual.
2. The directed network optimization method of screening robust influence nodes of claim 1, wherein the cross-over operation comprises:
Exchanging nodes of two random individuals in the first population based on preset gene points;
and after the crossing operation, when a repeated node exists in the individual of the exchanged node, acquiring a new node from the target directed network to replace the repeated node.
3. The directed network optimization method of claim 1, wherein the mutation operation comprises:
obtaining a new node from the target directed network to replace a random node of any individual in the second population;
wherein the new node does not repeat with any node in the individual of replaced nodes.
4. The directed network optimization method for screening robust influence nodes according to claim 1, wherein the searching the third population, updating the individuals in the third population, further performing fitness evaluation on the updated individuals, and obtaining a fourth population based on the evaluation result, includes:
performing global search operation on first individuals in the third population, performing node replacement on the first individuals in the third population through target nodes in the target directed network, further performing fitness evaluation on the first individuals before and after the node replacement respectively, and when the robust influence of the first individuals after the node replacement is improved, reserving the first individuals after the node replacement;
And/or performing neighborhood searching operation on the second individuals in the third population, performing node replacement on the second individuals through traversing and searching neighborhood nodes of all nodes in the second individuals in the third population, performing fitness evaluation on the second individuals before and after the node replacement by using random neighborhood nodes, and retaining the second individuals after the node replacement when the robustness influence of the second individuals after the node replacement is improved;
and/or performing random search operation on a third individual in the third population, performing node replacement on the third individual by randomly acquiring a new node from the target directed network, further performing fitness evaluation on the third individual before and after the node replacement, and when the robust influence of the third individual after the node replacement is improved, reserving the third individual after the node replacement;
and obtaining a fourth population according to the individuals of the third population remained after node replacement.
5. The directed network optimization method for screening robust influence nodes according to claim 1, wherein said performing said fitness evaluation on all individuals of said fourth population, and further selecting individuals from said fourth population for population iteration according to the evaluation result, comprises:
Carrying out the fitness evaluation on all individuals of the fourth population to obtain the robust influence of all the individuals of the fourth population;
and sorting all individuals of the fourth population according to the level of the robust influence, screening the individuals of the new population of the sorted fourth population through sequential sorting selection and roulette selection, and further carrying out population iteration according to the screened individuals of the new population.
6. A directed network optimization method for screening robust influence nodes according to any of claims 1 to 5, wherein said fitness evaluation comprises:
continuously simulating network attack on the target directed network, removing the attacked nodes in the target directed network one by one, and further determining the robust influence of the evaluated individual according to the influence level of the evaluated individual in the network structure of each stage of the removed nodes of the target directed network and the control link number of the network structure of the target directed network in each stage of the removed nodes.
7. The directed network optimization method of screening robust influence nodes of claim 6, wherein said determining the robust influence of the evaluated individual based on the influence level of the evaluated individual in the network structure of the target directed network at each stage of the removal node and the number of control links of the network structure of the target directed network at each stage of the removal node comprises:
Determining the influence level by the difference between the sum of the influence ranges of the evaluated individuals in the network structure of each stage of the removed node and the overlapping influence;
determining a state value of the evaluated individual at each stage of the removal node of the target directional network according to the ratio of the influence level to the control link number;
and carrying out accumulation normalization processing on the state values of the evaluated individuals at all stages of the removal node of the target directed network to obtain the robust influence of the evaluated individuals on the target directed network.
8. A directed network optimization system for screening robust influence nodes, comprising:
the first module is used for carrying out node random selection on the target directed network to obtain a seed node set, and repeating the node random selection to obtain an initial population; wherein said initial population comprises a plurality of individuals, each of said individuals characterizing a different set of said seed nodes;
the second module is used for taking the initial population as a first population, performing cross operation on the first population, and obtaining a second population;
the third module is used for carrying out mutation operation on the second population to obtain a third population;
A fourth module, configured to perform a search operation on the third population, update individuals in the third population, further perform fitness evaluation on the updated individuals, and obtain a fourth population based on an evaluation result;
a fifth module, configured to perform the fitness evaluation on all the individuals in the fourth population, further select individuals from the fourth population according to an evaluation result, perform population iteration, use the iterated population as a first population, and then return to the second module to perform a cross operation on the first population, so as to obtain a second population, until reaching a preset iteration number, and use the population after the last iteration as a target population;
a sixth module, configured to perform the fitness evaluation on all the individuals in the target population, and obtain, as a target seed node set, an individual with the greatest robustness in the target population;
wherein the first population, the second population, the third population, the fourth population, and the target population each comprise a plurality of individuals, each individual comprising a number of nodes in the target-directed network; the fitness assessment includes an assessment of the robust impact of the assessed individual.
9. An electronic device comprising a processor and a memory;
the memory is used for storing programs;
the processor executing the program implements the method of any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the storage medium stores a program that is executed by a processor to implement the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311003446.5A CN117155786B (en) | 2023-08-09 | Directed network optimization method and system for screening robust influence nodes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311003446.5A CN117155786B (en) | 2023-08-09 | Directed network optimization method and system for screening robust influence nodes |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117155786A true CN117155786A (en) | 2023-12-01 |
| CN117155786B CN117155786B (en) | 2024-07-16 |
Family
ID=
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107517201A (en) * | 2017-07-28 | 2017-12-26 | 北京航空航天大学 | A kind of network vulnerability discrimination method removed based on sequential |
| CN110402567A (en) * | 2016-12-29 | 2019-11-01 | 华为技术有限公司 | Central caching is based in network centered on information |
| CN111126758A (en) * | 2019-11-15 | 2020-05-08 | 中南大学 | Academic team influence propagation prediction method, device and storage medium |
| CN115277115A (en) * | 2022-07-06 | 2022-11-01 | 中山大学 | Method and system for solving robust information propagation problem on network |
| CN115577283A (en) * | 2022-09-16 | 2023-01-06 | ***股份有限公司 | Entity classification method and device, electronic equipment and storage medium |
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110402567A (en) * | 2016-12-29 | 2019-11-01 | 华为技术有限公司 | Central caching is based in network centered on information |
| CN107517201A (en) * | 2017-07-28 | 2017-12-26 | 北京航空航天大学 | A kind of network vulnerability discrimination method removed based on sequential |
| CN111126758A (en) * | 2019-11-15 | 2020-05-08 | 中南大学 | Academic team influence propagation prediction method, device and storage medium |
| CN115277115A (en) * | 2022-07-06 | 2022-11-01 | 中山大学 | Method and system for solving robust information propagation problem on network |
| CN115577283A (en) * | 2022-09-16 | 2023-01-06 | ***股份有限公司 | Entity classification method and device, electronic equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 宋甲秀;杨晓翠;张曦煌;: "复杂网络中Top-k影响力节点的识别算法", 计算机科学与探索, no. 06 * |
| 王帅等: "一种针对网络结构破坏下鲁棒影响力最大化问题的 Memetic 算法", 《计算机学报》, vol. 44, no. 6, pages 1153 - 1167 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shrikumar et al. | Technical note on transcription factor motif discovery from importance scores (TF-MoDISco) version 0.5. 6.5 | |
| Lichtnwalter et al. | Link prediction: fair and effective evaluation | |
| Ben-Naim et al. | Complex networks | |
| McDiarmid et al. | Counting phylogenetic networks | |
| Ljubic | Exact and memetic algorithms for two network design problems | |
| García-Martínez et al. | A simulated annealing method based on a specialised evolutionary algorithm | |
| Poladian et al. | Multi-objective evolutionary algorithms and phylogenetic inference with multiple data sets | |
| Sun | A study of solving traveling salesman problem with genetic algorithm | |
| CN108204821A (en) | A kind of paths planning method and device | |
| Stripinis et al. | An extensive numerical benchmark study of deterministic vs. stochastic derivative-free global optimization algorithms | |
| CN117155786B (en) | Directed network optimization method and system for screening robust influence nodes | |
| CN117155786A (en) | Directed network optimization method and system for screening robust influence nodes | |
| Bara’a et al. | A new evolutionary multi-objective community mining algorithm for signed networks | |
| Consoli et al. | Heuristic approaches for the quartet method of hierarchical clustering | |
| Nikulin | Simulated annealing algorithm for the robust spanning tree problem | |
| Huebler et al. | Constructing semi-directed level-1 phylogenetic networks from quarnets | |
| CN113590912A (en) | Cross-social network alignment method fusing relative position and absolute degree distribution of nodes | |
| Hacid et al. | Incremental neighborhood graphs construction for multidimensional databases indexing | |
| CN111680196A (en) | Key node searching method based on bipartite graph butterfly structure | |
| Fernandes et al. | A multi-agent transgenetic algorithm for the bi-objective spanning tree problem | |
| Parsa et al. | An EDA-based community detection in complex networks | |
| Bütün et al. | A multi-objective genetic algorithm for community discovery | |
| Shang et al. | Evolutionary multi-objective overlapping community detection based on fusion of internal and external connectivity and correction of node intimacy | |
| CN116720293A (en) | Robust influence optimization method and device, electronic equipment and storage medium | |
| CN113487871B (en) | Rapid traffic distribution method, device and storage medium based on network aggregation strategy |
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 |