Deep-Learning-ToolBox-CNN-master
所属分类：模式识别(视觉/语音等)
开发工具：matlab
文件大小：57465KB
下载次数：33
上传日期：2017-11-12 12:41:24
上 传 者：
Oceanidiot
说明： 实现CNN的手写字体matlab程序，可调节激活函数，选择loss函数
(Realization of handwritten font matlab CNN procedures, can regulate the activation function, loss function)
文件列表:
CNN (0, 2015-05-08)
CNN\CNN_Test.m (2621, 2015-05-08)
CNN\MaxPooling.cpp (8899, 2015-05-08)
CNN\MaxPooling.mexw64 (13312, 2015-05-08)
CNN\ReLu.m (52, 2015-05-08)
CNN\cnnapplygrads.m (575, 2015-05-08)
CNN\cnnbp.m (4728, 2015-05-08)
CNN\cnnff.m (3653, 2015-05-08)
CNN\cnnnumgradcheck.m (3430, 2015-05-08)
CNN\cnnsetup.m (2283, 2015-05-08)
CNN\cnntest.m (417, 2015-05-08)
CNN\cnntrain.m (943, 2015-05-08)
CNN\expand.m (1958, 2015-05-08)
CNN\expansionData.m (295, 2015-05-08)
CNN\flipall.m (80, 2015-05-08)
CNN\loadMNISTImages.m (799, 2015-05-08)
CNN\loadMNISTLabels.m (706, 2015-05-08)
CNN\mnist_uint8.mat (14735220, 2015-05-08)
CNN\sigm.m (67, 2015-05-08)
CNN\softmax.m (256, 2015-05-08)
CNN\t10k-images.idx3-ubyte (7840016, 2015-05-08)
CNN\t10k-labels.idx1-ubyte (10008, 2015-05-08)
CNN\train-images.idx3-ubyte (47040016, 2015-05-08)
CNN\train-labels.idx1-ubyte (60008, 2015-05-08)
CNN\workspace_40.mat (33211883, 2015-05-08)
# Deep-Learning-ToolBox-CNN
Author: PIERRE_HAO
Data: 2015/05/01
An improved version for Deep Learning Toolbox CNN
1. Add ReLu function
2. Add Max pooling function
3. Add function to read MNIST which we can get from Yann LeCun Home Page
4. Pad size 28*28 to 32*32 with value 0
5. Other minor changes which are not important
6. Add softmax
7. Add stepsize for changing learning rate
@MASTERSTHESIS\{IMM2012-06284,
author = "R. B. Palm",
title = "Prediction as a candidate for learning deep hierarchical models of data",
year = "2012",
}
近期下载者：
相关文件：
收藏者：