Estimation of Distribution Algorithm using Two Kinds of Information f or Solving Knapsack Problem

摘要

Former information of probability model and inferior individuals was discarded in the research of estimation of distribution algorithm usually, but they may contain useful information. In this paper, the information of former probability is introduced to avoid premature convergence, and the information of inferior individuals is used to filter generated individuals. The algorithm is used to solve knapsack problem and simulated based on the widely used examples, the results demonstrate the effectiveness of proposed method, the utlilzation of former information of probability model and inferior individuals greatly improve the performance of algorithm.