基于输入扩展改进的BP网络及其在遥感图像分类中的应用

提出了一种基于输入模式扩展的神经网络改进方法,并和Levenberg-Marquardt优化的BP网络(LMBPN)进行了对比。通过二阶内积或切比雪夫多项式等非线性函数,把输入向量映射到更高维的模式空间,可以增强样本的可分性。Iris数据和遥感图像分类实验表明,输入模式扩展的神经网络改进方法可以进一步加快收敛速度,改进模式分类效果。

Input-Expansion-Based Improved Method for BP Network and Its Application for Remote Sensing Image Classification

The paper presents an input-expansion-based improved method for Neural Network,and make camparison for Levenberg-Marquardt optimized Back-Propagation Network(LMBPN).Using second-order inner product function or Chebyshev polynomial function to expand input vector of samples,which makes them mapped into a higher-dimensional pattern space and thus leads to the sample sets more easily separable.Iris data and remote sensing data classification experiments show that input-expansion method may efficiently enhance the convergence velocity and classification accuracy for neural network.