A Novel Chaotic Neural Network with Stochastic Noise and Heuristic Mechanism for Minimum Vertex Cover Problem

In this paper, we propose a novel chaotic neural network embedded with stochastic simulated annealing noise and a heuristic mechanism to solve minimum vertex cover problem. The proposed network can make a global search with the affection of stochastic noise and obtain a chaotic search by the chaotic dynamics. The stochastic noise with simulated annealing is able to find a global optimum solution if the annealing process is carried out sufficiently slowly. For increasing the network convergence speed and degree, a heuristic mechanism on vertex degree is introduced to modify the convergence trend. The proposed network is tested on a large number of random graphs. The simulation results show that the proposed algorithm is effective and better than some other works in solving minimum vertex cover problem.