摘要: In recent years it has been shown that an intelligent combination of metaheuristics with other optimization techniques can significantly improve over the application a pure metaheuristic. this paper, we combine evolutionary computation paradigm dynamic programming for to NP-hard k-cardinality tree problem. Given undirected graph G node and edge weights, problem consists finding in exactly k edges such sum weights is minimal. The genetic operators our algorithm are based on existing from literature optimal subtrees given tree. simulation results show able best known benchmark problems 60 cases.