A hybrid algorithm based on estimation of distribution algorithm is proposed to solve multidimensional knapsack
problem(MKP). In the algorithm, the probabilistic model is built with the superior population and new individuals are
generated based on probability model. Meanwhile, a repair operator based on MKP specific knowledge is proposed to
effectively repair infeasible solutions generated by sampling. In addition, an adaptive local search is designed to enhance the
exploitation ability. Simulation results based on benchmark testing problems and comparisons demonstrate the effectiveness
and robustness of the proposed hybrid algorithm.