一种改进的LeNet-5卷积神经网络算法

An Improved LeNet-5 Convolutional Neural Network Algorithm

LeNet-5卷积神经网络是一种手写数字识别的模型。其通过对输入数据的卷积、激活和池化操作,提取特征值,以达到图像分类。文章在卷积神经网络优化原则的基础上,提出了改良的LeNet-5卷积神经网络模型,该模型优化了激活函数,使用全连接层代替了最后的卷积层,提高了计算速度和稳定性。选取MNIST作为数据集进行了实验,该改进模型的实验结果表明,具其备更好的辨别精度。

Lenet-5 convolutional neural network is a model of handwritten digit recognition.Through convolution,activation and pooling of input data,the feature values are extracted to achieve image classification.Based on the principle of convolution neural network optimization,an improved lenet-5 convolutional neural network model is proposed in this paper.The model optimizes the activation function and the final convolution layer is replaced by the full connection layer,which improves the calculation speed and stability.MNIST was selected as the data set for the experiment,and the experimental results of the improved model showed that it had better discrimination accuracy.