基于主成分分析和神经网络的目标识别

采用主成分分析法将多项识别属性进行综合聚集,减少关联属性对识别的干扰,达到属性约简的目的。利用神经网络进行目标识别,通过优化训练策略,可以提高网络的收敛速度和泛化能力。主成分分析法与神经网络结合既能高效识别已知样本,又能对未知样本具有很强的自学与适应能力,从而有效地对海战场目标进行识别。通过对比仿真,证明了算法的有效性。

Target Recognition Based on Principal Component Analysis and Neural Networks

The principal component analysis(PCA) is used to aggregate the recognition attribute in order to decrease the association of each attribute and reduce the attribute.The neural networks is used to recognize the target.The use of optimizing policy can improve the constringency speed and the generalization ability of the neural networks.The combined method of principal component analysis and neural networks not only can recognize the target in high efficiency,but also can have the ability of self-study and adapting which can recognize the target in naval battlefield.A simulation is given to prove the efficiency of this algorithmic.