论文部分内容阅读
分析了文本分类系统的一般模型及现有技术,在应用了核主成分分析的特征降维方法进行处理后,提出了一种基于样本中心的径向基(RBF)神经网络文本分类算法,并且引入了聚类算法的核心思想,来改进误差反向传播(BP)神经网络分类算法收敛速度较慢的缺点。实验结果表明,RBF网络与BP网络相比,具有较高的运算速度和较强的非线性映射能力,在收敛速度和准确程度上也有更好的分类效果。
After analyzing the general model of text classification system and the existing techniques, a feature-based radial basis-based (RBF) neural network text classification algorithm based on sample center is proposed after feature dimensionality reduction method using kernel principal component analysis The core idea of clustering algorithm is introduced to improve the shortcoming of convergence error of BP neural network classification algorithm. Experimental results show that compared with BP network, RBF network has higher computational speed and stronger nonlinear mapping ability, and also has better classification results in terms of convergence speed and accuracy.