Convolutional-Neural-Network-master
所属分类:其他
开发工具:matlab
文件大小:15632KB
下载次数:20
上传日期:2018-10-30 21:06:27
上 传 者:
yaminglu
说明: 在MATLAB中利用卷积神经网络实现手写数字的识别
(Convolutional-Neural-Network-master)
文件列表:
Example.m (1312, 2015-06-19)
InitializeParameters.m (1811, 2015-06-19)
cnn.mat (4510134, 2015-06-19)
cnnConvolve4D.m (930, 2015-06-19)
cnnPool.m (1538, 2015-06-19)
cnnTest.m (850, 2015-06-19)
cnnTrain.m (7239, 2015-06-19)
common (0, 2015-06-19)
common\loadMNISTImages.m (811, 2015-06-19)
common\loadMNISTLabels.m (516, 2015-06-19)
common\t10k-images-idx3-ubyte (7840016, 2015-06-19)
common\t10k-labels-idx1-ubyte (10008, 2015-06-19)
common\train-images-idx3-ubyte (47040016, 2015-06-19)
common\train-labels-idx1-ubyte (60008, 2015-06-19)
# Convolutional-Neural-Network
This is a matlab implementation of CNN on MNIST
It can have as many layers as you want, an example of setting structure of a neural network is as below:
cnn.layers = {
struct('type', 'c', 'numFilters', 6, 'filterDim', 5)
struct('type', 'p', 'poolDim', 2)
struct('type', 'c', 'numFilters', 8, 'filterDim', 5)
struct('type', 'p', 'poolDim', 2) %subsampling layer
};
The above code sets up a two-layer convolutional neural network.
To try the demo, just run Example.m.
近期下载者:
相关文件:
收藏者: