Weak Structure Balance Analysis Of Signed Networks Based On Evolutionary Algorithms

In the real life,the connections between objects can be abstracted into complex networks.Such as social networks,international relations networks,power networks,biological networks,and so on.In general,a graph is used to represent a complex network.The nodes in the graph represent the objects in the network,and the edges represent the connection between objects.As the study progressed,researchers gradually found that there are not only positive connection,but also negative correlation between objects.For example,in social networks,people are not only friends,but also enemies.In biological networks,the interaction between cells is not only facilitated but also inhibited.So researchers call such networks as signed networks.In the process of the dynamic evolution,the signed networks will develop slowly to the stable structure over time.The structure balance analysis is an important research direction,which is significance to understand the feature of network global structure.Therefore,it is very important to study the structure balance property of signed network.For the real signed network,the strong structure balance is too demanding,and it is difficult to achieve strong structural balance.Therefore,this paper uses the weak structure balance theorem to solve the network structure balance.As we all know,the structure balance property of signed network is a NP hard problem,we can't solve it directly.Based on this,evolutionary algorithms are used to solve the problem.Based on the current research results,the mainly works in this paper are introduced as follows:1.Proposing the energy function of weak structural balance theorem.The minimum value of the energy function is the weak unbalance degree of the network.Basised on the energy function of the structural balance theorem,We add the kronecker function to transform it into the energy function of the weak structural balance theorem.2.Using the evolutionary algorithms to solve the weak structural balance problem of the signed network,the WSB-EA series algorithms are proposed.We chose suitable initializers,selection,crossover and mutation methods to achieve it.3.Using the incremental calculation method to reduce the time complexity of the algorithm.After each crossover and mutation operation,it is necessary to recalculate the adaptive value.That is the reason why the algorithm runs too slowly.Therefore,this paper proposes an incremental calculation method to reduce the difficulty of calculating the adaptive value and accelerate the speed of the algorithm.4.Using compression coding reduces the individual length and reduces the spatial complexity of the algorithm.According to the principle proposed in this paper,an indirect coding method is designed to shorten the length of the individual,thus reducing the memory footprint and speeding up the algorithm.