摘要
介绍了RBF神经网络在人脸识别中的应用;通过理论分析和实验效果突出了RBF神经网络在机器学习中的优势。实验结果证明:RBF神经网络具有运算速度快、识别率高、算法简单等特点。在训练样本减少的情况下,该学习机的分类性能没有明显退化。因此,RBF神经网络是一种性能优异的学习机。
This paper introduces RBF neural network used in face recognition. Through theoretical analysis and experimental results make the advantage of RBF neural network outstanding in machine learning. Experimental results prove that RBF neural network has some excellent features such as high speed for computing, high correct rate for classification, rather simple algorithm etc. Furthermore, the performance of classification doesn＇t degenerate obviously when training sample decrease. Therefore RBF neural network is an excellent learning machine.
出处
《光学仪器》
2008年第2期31-33,共3页
Optical Instruments
