利用汉明距离优选神经网络学习样本

Selecting of Learning Samples Based on Hamming Distance

鉴于学习样本对神经网络模型的模式识别性能有很大的影响 ,提出学习样本的选择应与识别模型所利用的特性相结合 ,并利用汉明 (Hamming)距离对用于旋转不变识别的级联模型的学习样本进行优选 ,计算机对三个很相似的飞机模型进行识别 ,识别结果表明对学习样本进行有效的选择不仅可以减少系统的学习训练时间而且可以提高模型的识别能力。

Learning samples significantly affect the recognition ability of neuron network models. One of selecting rules of learning samples is proposed according to the principle of the pattern recognition model. A method of selecting learning samples based on Hamming distance used in the cascade neuron network model for rotation invariance recognition is analyzed. The results of the computer recognition show that the effective selection of the learning samples can not only reduce the training time but also improve the recognition ability of the model.