基于PSO的BP训练算法论文：在BP训练算法中，关于变权值、学习速率、步长的问题已被广泛地研究，几种基于启发式改进的技术也表明具有改善训练时间以及避免陷入局部最小的明显效果。这里BP训练过程由基于PSO同时优化log—Sigmoid函数与网络权值的新算 法(PSO。GainBP)实现。实验结果表明，PSO—GainBP比传统基于PSO的BP算法在网络训练方面具有更好的性能。-PSO based on the BP training algorithm Papers: In the BP tr

利用bp神经网络对遥感图像进行分类，输入样本值后，根据样本值对遥感图像不同的地物进行分类，分类后计算每种地物所占面积-The use of bp neural network classification of remote sensing images, enter a sample value, based on the sample value of the different features of remote sensing image classification, classif

BP神经网络的整个训练过程,中介真值程度度量，基于中介真值程度度量的图像分割MIMO OFDM matlab仿真。- The entire training process BP neural network, The true extent of the value of the intermediary measure, measure the true extent of the agency based on the value of image segmentation MIMO OF