基于思维进化算法的径向基函数神经网络结构优化

为了解决一类径向基函数的结构优化问题，该文在基本思维进化计算框架的基础上，提出了一种有效的混合优化策略。在优化过程中充分利用样本的信息，同时借鉴信息矩阵的思想，提出了利用信息矩阵进行信息抽取和信息积累的方法，并设计了有效的趋同、异化算子与个体之间学习的具体过程，使结构和参数同时得到了优化。仿真研究表明，该算法是快速有效的，并能保证网络具有较好的泛化能力。

RBF Neural Network Structure Optimization Based on MEC Algorithm

To solve the structure optimization of a class of radial basis function neural network, based on fundamental mind evolutionary computation (MEC), an effective hybrid strategy is presented. The swatch information is considered enough in evolution process, use for reference the idea of information matrix, a method of information extract and information accumulation based on information matrix is presented, at the same time, the specific procedure of learning between individuals and effective operators of similartaxis and dissimilation are designed, the method is able to determine the structure and parameters. Simulation results show the method is not only very feasible and effective but also can improve the generalization ability.