A warning propagation algorithm to solve the double-objective minimum spanning tree

The double-objective minimum spanning tree problem is an NP-hard problem with important application values in numerous fields such as optical cable communication and intelligent control. An undirected graph is transformed into a factor graph using the restricted Boltzmann machine, and the double-objective minimum spanning tree problem in the undirected graph is mapped into the corresponding problem in the factor graph. To construct the adjacency matrix, several random numbers generated from random number seeds are selected. Moreover, a corresponding undirected graph example is generated. Numerical experimental results indicate that the proposed algorithm is better than other similar algorithms.