基于遗传算法的RBM优化设计

RBM Optimization Based on Genetic Algorithms

为了有效解决受限玻尔兹曼机在设计时没有规律遵循并很难保证网络最优化的问题,提出一种基于遗传算法的RBM辅助优化设计方法 (Genetic Algorithm-Restricted Boltzmann Machine,GA-RBM),完成了RBM模型结构和权值的全局搜索.针对RBM特点,设计RBM模型个体编码方式和适应度函数,实现了通过遗传算法对可见单元维度的优化和隐单元个数的选择.通过MNIST实验证明,相较于其他常规的数据降维方式,该方法不仅可以降低可见单元维数,而且能够有效提高RBM特征提取性能,达到了通过遗传算法实现RBM模型优化设计的目的.

To resolve the problem of no guidance about how to set values of numerical meta-parameters parameters and difficulty to achieve optimization of Restricted Boltzmann Machines,genetic algorithms are used to develop an automatic optimizing method named GA-RBM(Genetic Algorithm-Restricted Boltzmann Machine)for Restricted Boltzmann Machines′aided design.Based on the features of Restricted Boltzmann Machines and evaluation function,agenetic algorithm is designed and realizes a global search of satisfied structure and network′s weights to determine the number of visible units and hidden units,and also initial weights.Firstly we introduced RBM parameters affect the ability of the model to extract features,studied the genetic algorithm and a combination of RBM training,then proposed RBM model of individual coding and fitness function is designed for RBM training can effectively achieve through genetic algorithm simultaneously search for RBM network structure and weights.The experiments were conducted on MNIST digits handwritten datasets.The results proved that this optimization reduces the dimension of visible units;improves the performance of feature extracted by Restricted Boltzmann Machines.The network optimized has good generalization performance and meets the demand of Restricted Boltzmann Machines′aided design.