DeepLearnToolbox使用总结

DeepLearnToolbox

A Matlab toolbox for Deep Learning.

Deep Learning is a new subfield of machine learning that focuses on learning deep hierarchical models of data. It is inspired by the human brain's apparent deep (layered, hierarchical) architecture. A good overview of the theory of Deep Learning theory is Learning Deep Architectures for AI

Directories included in the toolbox

NN/ - A library for Feedforward Backpropagation Neural Networks

CNN/ - A library for Convolutional Neural Networks

DBN/ - A library for Deep Belief Networks

SAE/ - A library for Stacked Auto-Encoders

CAE/ - A library for Convolutional Auto-Encoders

util/ - Utility functions used by the libraries

data/ - Data used by the examples

tests/ - unit tests to verify toolbox is working

For references on each library check REFS.md

Setup

Download.
addpath(genpath('DeepLearnToolbox'));

Windows下把文件夹加入 path 即可

%LiFeiteng 

path = pwd;
files = dir(path);

for i = 3:length(files)
    
    if files(i).isdir        
        file = files(i).name;       
        addpath([path '/' file])
        disp(['add ' file ' to path!'])
    end   
    
end

我不打算解析代码，想从代码里面学算法是stupid的；有相应的论文，readlist，talk等可以去学习。

DeepLearnToolbox单隐藏层NN的优化策略：mini-Batch SGD

function [nn, L]  = nntrain(nn, train_x, train_y, opts, val_x, val_y)
%NNTRAIN trains a neural net
% [nn, L] = nnff(nn, x, y, opts) trains the neural network nn with input x and
% output y for opts.numepochs epochs, with minibatches of size
% opts.batchsize. Returns a neural network nn with updated activations,
% errors, weights and biases, (nn.a, nn.e, nn.W, nn.b) and L, the sum
% squared error for each training minibatch.

assert(isfloat(train_x), 'train_x must be a float');
assert(nargin == 4 || nargin == 6,'number ofinput arguments must be 4 or 6')

loss.train.e               = [];
loss.train.e_frac          = [];
loss.val.e                 = [];
loss.val.e_frac            = [];
opts.validation = 0;
if nargin == 6
    opts.validation = 1;
end

fhandle = [];
if isfield(opts,'plot') && opts.plot == 1
    fhandle = figure();
end

m = size(train_x, 1);

batchsize = opts.batchsize;
numepochs = opts.numepochs;

numbatches = m / batchsize;

assert(rem(numbatches, 1) == 0, 'numbatches must be a integer');

L = zeros(numepochs*numbatches,1);
n = 1;
for i = 1 : numepochs
    tic;
    
    kk = randperm(m);
    for l = 1 : numbatches
        batch_x = train_x(kk((l - 1) * batchsize + 1 : l * batchsize), :);
        
        %Add noise to input (for use in denoising autoencoder)
        if(nn.inputZeroMaskedFraction ~= 0)
            batch_x = batch_x.*(rand(size(batch_x))>nn.inputZeroMaskedFraction);
        end
        
        batch_y = train_y(kk((l - 1) * batchsize + 1 : l * batchsize), :);
        
        nn = nnff(nn, batch_x, batch_y);
        nn = nnbp(nn);
        nn = nnapplygrads(nn);
        
        L(n) = nn.L;
        
        n = n + 1;
    end
    
    t = toc;
    
    if ishandle(fhandle)
        if opts.validation == 1
            loss = nneval(nn, loss, train_x, train_y, val_x, val_y);
        else
            loss = nneval(nn, loss, train_x, train_y);
        end
        nnupdatefigures(nn, fhandle, loss, opts, i);
    end
        
    disp(['epoch ' num2str(i) '/' num2str(opts.numepochs) '. Took ' num2str(t) ' seconds' '. Mean squared error on training set is ' num2str(mean(L((n-numbatches):(n-1))))]);
    nn.learningRate = nn.learningRate * nn.scaling_learningRate;
end
end

1.不管是在 nntrain、 nnbp还是nnapplygrads中我都没看到对算法收敛性的判断，

而且在实测的过程中有观察到 epoch过程中 mean-squared-error有下降-上升-下降的走势——微小抖动在SGD中算是正常

多数还都是在下降（epoch我一般设为 10-40，这个值可能偏小；Hinton 06 science的文章代码记得epoch了200次，我跑了3天也没跑完）

在SAE/CNN等中也没看到收敛性的判断。

2.CAE 没有完成

3.dropout的优化策略也可以选择

我测试了 SAE CNN等，多几次epoch(20-30)，在MNIST上正确率在 97%+的样子。

其实cost-function 可以有不同的选择，如果使用 UFLDL的优化方式(固定的优化方法，传入cost-function的函数句柄)，在更改cost-function上会更自由。

可以改进的地方：

1. mini-Bathch SGD算法增加收敛性判断

2.增加 L-BFGS/CG等优化算法

3.完善CAE等

4.增加min KL-熵的 Sparse Autoencoder等

5.优化算法增加对不同cost-function的支持