【摘要】 为解决复杂的非线性可分问题中前馈式神经网络收敛缓慢的问题 ,提出一种新的神经网络模型———自适应特征空间扩张神经网络 .该模型在标准前馈式神经网络的前端增加一个自适应特征空间扩张层 ,自适应地增强原始模式的表达 ,将原来的非线性可分问题转换成线性可分问题或者减小其非线性程度 ,从而加快网络的收敛速度 .一系列的实验 ,特别是双螺旋线实验表明 ,该网络模型不仅具有很快的收敛速度 ,而且具有很好的泛化能力更多还原

【Abstract】 Most of the pattern recognition problems are nonlinear separable. For these complicated nonlinear separable problems, the feedforward neural network usually converges slowly. In this paper, we propose a new kind of neural network model named Self adaptive Feature Space Expanding Neural Network to overcome this problem. In this model, a self adaptive feature space expanding layer is added in front of the feedforward neural network to enhance the description of the original pattern, thus the nonlinear separable problem can be transformed into linear one or less nonlinear one, and our new model converges faster than the traditional feedforward neural network. A series of experiments, especially the two spiral problem indicate that our new model converges very fast, and has good generalization.更多还原