• 自编码算法与稀疏性


    前言

    看完神经网络及BP算法介绍后,这里做一个小实验,内容是来自斯坦福ULIDL教程,实现图像的压缩表示,模型是用神经网络模型,训练方法是BP后向传播算法。

    理论

           在有监督学习中,训练样本是具有标签的,一般神经网络是有监督的学习方法。我们这里要讲的是自编码神经网络,这是一种无监督的学习方法,它是让输出值等于自身来实现的。
         
          从图中可以看到,神经网络模型只有一层隐含层,输出层跟输入层的神经单元个数是一样的。如果隐含层单元个数比输入层少的话,我们用这个模型学到的是输入数据的压缩表示,相当于对输入数据进行降维(这是一种非线性的降维方法)。实际上,如果隐含层单元个数比输入层多,我们可以让隐含层的大部分单元激活值接近0,就是让它们稀疏,这样学到的也是压缩表示。我们模型要使得输出层跟输入层一样,就是隐含层要能够重建出跟输入层一样的输出层,这样我们学到的压缩表示才是有意义的。
         回忆下之前介绍过的损失函数:
       
        在这里,y是输出层,跟输入层是一样的。
        自编码神经网络还增加了稀疏性惩罚一项。它是对隐含层进行了稀疏性的约束,即使得隐含层大部分值都处于非active状态。定义隐含层节点j的稀疏程度为
       
         上式是对整个样本求隐含层节点j的平均值,如果是所有隐含层节点,那么就组成一个向量。
         我们要设置期望隐含层稀疏性的程度,假设为,因此我们希望对于所有的节点j
         那怎么衡量实际跟期望的差别呢?
         
         实际上是关于伯努利变量p与q的KL离散度(参考我之前写的关于信息熵的博客)。
        此时损失函数为
        
        由于加了稀疏项损失函数,对第二层节点求残差时公式变为
       
     

    实验

           实验教程是在Exercise:Sparse Autoencoder,要实现的文件是sampleIMAGES.m, sparseAutoencoderCost.m,computeNumericalGradient.m
         
         实验步骤:
    1. 生成训练集
    2. 稀疏自编码目标函数
    3. 梯度校验
    4. 训练稀疏自编码
    5. 可视化
           最后一步可视化是,把x用图像表示出来的。
      
          
        代码如下:
    sampleIMAGES.m
    function patches = sampleIMAGES()
    % sampleIMAGES
    % Returns 10000 patches for training
    
    
    load IMAGES;    % load images from disk 
    
    
    patchsize = 8;  % we'll use 8x8 patches 
    numpatches = 10000;
    
    
    % Initialize patches with zeros.  Your code will fill in this matrix--one
    % column per patch, 10000 columns. 
    patches = zeros(patchsize*patchsize, numpatches);
    
    
    %% ---------- YOUR CODE HERE --------------------------------------
    %  Instructions: Fill in the variable called "patches" using data 
    %  from IMAGES.  
    %  
    %  IMAGES is a 3D array containing 10 images
    %  For instance, IMAGES(:,:,6) is a 512x512 array containing the 6th image,
    %  and you can type "imagesc(IMAGES(:,:,6)), colormap gray;" to visualize
    %  it. (The contrast on these images look a bit off because they have
    %  been preprocessed using using "whitening."  See the lecture notes for
    %  more details.) As a second example, IMAGES(21:30,21:30,1) is an image
    %  patch corresponding to the pixels in the block (21,21) to (30,30) of
    %  Image 1
    [m,n,num] = size(IMAGES);
    
    
    for i=1:numpatches
        j = randi(num);
        bx = randi(m-patchsize+1);
        by = randi(n-patchsize+1);
        block = IMAGES(bx:bx+patchsize-1,by:by+patchsize-1,j);
        
        patches(:,i) = block(:);
    end
    
    
    
    
    
    
    %% ---------------------------------------------------------------
    % For the autoencoder to work well we need to normalize the data
    % Specifically, since the output of the network is bounded between [0,1]
    % (due to the sigmoid activation function), we have to make sure 
    % the range of pixel values is also bounded between [0,1]
    patches = normalizeData(patches);
    
    
    end
    
    
    
    
    %% ---------------------------------------------------------------
    function patches = normalizeData(patches)
    
    
    % Squash data to [0.1, 0.9] since we use sigmoid as the activation
    % function in the output layer
    
    
    % Remove DC (mean of images). 
    patches = bsxfun(@minus, patches, mean(patches));
    
    
    % Truncate to +/-3 standard deviations and scale to -1 to 1
    pstd = 3 * std(patches(:));
    patches = max(min(patches, pstd), -pstd) / pstd;
    
    
    % Rescale from [-1,1] to [0.1,0.9]
    patches = (patches + 1) * 0.4 + 0.1;
    
    
    end
    
     
    SparseAutoencoderCost.m
    function [cost,grad] = sparseAutoencoderCost(theta, visibleSize, hiddenSize, ...
                                                 lambda, sparsityParam, beta, data)
    
    % visibleSize: the number of input units (probably 64) 
    % hiddenSize: the number of hidden units (probably 25) 
    % lambda: weight decay parameter
    % sparsityParam: The desired average activation for the hidden units (denoted in the lecture
    %                           notes by the greek alphabet rho, which looks like a lower-case "p").
    % beta: weight of sparsity penalty term
    % data: Our 64x10000 matrix containing the training data.  So, data(:,i) is the i-th training example. 
      
    % The input theta is a vector (because minFunc expects the parameters to be a vector). 
    % We first convert theta to the (W1, W2, b1, b2) matrix/vector format, so that this 
    % follows the notation convention of the lecture notes. 
    
    W1 = reshape(theta(1:hiddenSize*visibleSize), hiddenSize, visibleSize);
    W2 = reshape(theta(hiddenSize*visibleSize+1:2*hiddenSize*visibleSize), visibleSize, hiddenSize);
    b1 = theta(2*hiddenSize*visibleSize+1:2*hiddenSize*visibleSize+hiddenSize);
    b2 = theta(2*hiddenSize*visibleSize+hiddenSize+1:end);
    
    % Cost and gradient variables (your code needs to compute these values). 
    % Here, we initialize them to zeros. 
    cost = 0;
    W1grad = zeros(size(W1)); 
    W2grad = zeros(size(W2));
    b1grad = zeros(size(b1)); 
    b2grad = zeros(size(b2));
    
    %% ---------- YOUR CODE HERE --------------------------------------
    %  Instructions: Compute the cost/optimization objective J_sparse(W,b) for the Sparse Autoencoder,
    %                and the corresponding gradients W1grad, W2grad, b1grad, b2grad.
    %
    % W1grad, W2grad, b1grad and b2grad should be computed using backpropagation.
    % Note that W1grad has the same dimensions as W1, b1grad has the same dimensions
    % as b1, etc.  Your code should set W1grad to be the partial derivative of J_sparse(W,b) with
    % respect to W1.  I.e., W1grad(i,j) should be the partial derivative of J_sparse(W,b) 
    % with respect to the input parameter W1(i,j).  Thus, W1grad should be equal to the term 
    % [(1/m) Delta W^{(1)} + lambda W^{(1)}] in the last block of pseudo-code in Section 2.2 
    % of the lecture notes (and similarly for W2grad, b1grad, b2grad).
    % 
    % Stated differently, if we were using batch gradient descent to optimize the parameters,
    % the gradient descent update to W1 would be W1 := W1 - alpha * W1grad, and similarly for W2, b1, b2. 
    % 
    
    %矩阵向量化形式实现,速度比不用向量快得多
    Jocst = 0; %平方误差
    Jweight = 0; %规则项惩罚
    Jsparse = 0; %稀疏性惩罚
    [n, m] = size(data); %m为样本数,这里是10000,n为样本维数,这里是64
    
    %feedforward前向算法计算隐含层和输出层的每个节点的z值(线性组合值)和a值(激活值)
    %data每一列是一个样本,
    z2 = W1*data + repmat(b1,1,m); %W1*data的每一列是每个样本的经过权重W1到隐含层的线性组合值,repmat把列向量b1扩充成m列b1组成的矩阵
    a2 = sigmoid(z2);
    z3 = W2*a2 + repmat(b2,1,m);
    a3 = sigmoid(z3);
    
    %计算预测结果与理想结果的平均误差
    Jcost = (0.5/m)*sum(sum((a3-data).^2));
    %计算权重惩罚项
    Jweight = (1/2)*(sum(sum(W1.^2))+sum(sum(W2.^2)));
    %计算稀疏性惩罚项
    rho_hat = (1/m)*sum(a2,2);
    Jsparse = sum(sparsityParam.*log(sparsityParam./rho_hat)+(1-sparsityParam).*log((1-sparsityParam)./(1-rho_hat)));
    
    %计算总损失函数
    cost = Jcost + lambda*Jweight + beta*Jsparse;
    
    %反向传播求误差值
    delta3 = -(data-a3).*fprime(a3); %每一列是一个样本对应的误差
    sterm = beta*(-sparsityParam./rho_hat+(1-sparsityParam)./(1-rho_hat)); 
    delta2 = (W2'*delta3 + repmat(sterm,1,m)).*fprime(a2);
    
    %计算梯度
    W2grad = delta3*a2';
    W1grad = delta2*data';
    W2grad = W2grad/m + lambda*W2;
    W1grad = W1grad/m + lambda*W1;
    b2grad = sum(delta3,2)/m; %因为对b的偏导是个向量,这里要把delta3的每一列加起来
    b1grad = sum(delta2,2)/m;
    
    %%----------------------------------
    % %对每个样本进行计算, non-vectorial implementation
    % [n m] = size(data);
    % a2 = zeros(hiddenSize,m);
    % a3 = zeros(visibleSize,m);
    % Jcost = 0;    %平方误差项
    % rho_hat = zeros(hiddenSize,1);   %隐含层每个节点的平均激活度
    % Jweight = 0;  %权重衰减项   
    % Jsparse = 0;   % 稀疏项代价
    % 
    % for i=1:m
    %     %feedforward向前转播
    %     z2(:,i) = W1*data(:,i)+b1;
    %     a2(:,i) = sigmoid(z2(:,i));
    %     z3(:,i) = W2*a2(:,i)+b2;
    %     a3(:,i) = sigmoid(z3(:,i));
    %     Jcost = Jcost+sum((a3(:,i)-data(:,i)).*(a3(:,i)-data(:,i)));
    %     rho_hat = rho_hat+a2(:,i);  %累加样本隐含层的激活度
    % end
    % 
    % rho_hat = rho_hat/m; %计算平均激活度
    % Jsparse = sum(sparsityParam*log(sparsityParam./rho_hat) + (1-sparsityParam)*log((1-sparsityParam)./(1-rho_hat))); %计算稀疏代价
    % Jweight = sum(W1(:).*W1(:))+sum(W2(:).*W2(:));%计算权重衰减项
    % cost = Jcost/2/m + Jweight/2*lambda + beta*Jsparse; %计算总代价
    % 
    % for i=1:m
    %     %backpropogation向后传播
    %     delta3 = -(data(:,i)-a3(:,i)).*fprime(a3(:,i));
    %     delta2 = (W2'*delta3 +beta*(-sparsityParam./rho_hat+(1-sparsityParam)./(1-rho_hat))).*fprime(a2(:,i));
    % 
    %     W2grad = W2grad + delta3*a2(:,i)';
    %     W1grad = W1grad + delta2*data(:,i)';
    %     b2grad = b2grad + delta3;
    %     b1grad = b1grad + delta2;
    % end
    % %计算梯度
    % W1grad = W1grad/m + lambda*W1;
    % W2grad = W2grad/m + lambda*W2;
    % b1grad = b1grad/m;
    % b2grad = b2grad/m;
    
    % -------------------------------------------------------------------
    % After computing the cost and gradient, we will convert the gradients back
    % to a vector format (suitable for minFunc).  Specifically, we will unroll
    % your gradient matrices into a vector.
    grad = [W1grad(:) ; W2grad(:) ; b1grad(:) ; b2grad(:)];
    
    end
    
    %%      Implementation of derivation of f(z) 
    % f(z) = sigmoid(z) = 1./(1+exp(-z))
    % a = 1./(1+exp(-z))
    % delta(f) = a.*(1-a)
    function dz = fprime(a)
        dz = a.*(1-a);
    end
    %%
    %-------------------------------------------------------------------
    % Here's an implementation of the sigmoid function, which you may find useful
    % in your computation of the costs and the gradients.  This inputs a (row or
    % column) vector (say (z1, z2, z3)) and returns (f(z1), f(z2), f(z3)). 
    
    function sigm = sigmoid(x)
      
        sigm = 1 ./ (1 + exp(-x));
    end
    


    computeNumericalGradient.m
    function numgrad = computeNumericalGradient(J, theta)
    % numgrad = computeNumericalGradient(J, theta)
    % theta: a vector of parameters
    % J: a function that outputs a real-number. Calling y = J(theta) will return the
    % function value at theta. 
      
    % Initialize numgrad with zeros
    numgrad = zeros(size(theta));
    
    %% ---------- YOUR CODE HERE --------------------------------------
    % Instructions: 
    % Implement numerical gradient checking, and return the result in numgrad.  
    % (See Section 2.3 of the lecture notes.)
    % You should write code so that numgrad(i) is (the numerical approximation to) the 
    % partial derivative of J with respect to the i-th input argument, evaluated at theta.  
    % I.e., numgrad(i) should be the (approximately) the partial derivative of J with 
    % respect to theta(i).
    %                
    % Hint: You will probably want to compute the elements of numgrad one at a time. 
    EPSILON = 1e-4;
    
    for i=1:length(numgrad)
        theta1 = theta;
        theta1(i) = theta1(i)+EPSILON;
        theta2 = theta;
        theta2(i) = theta2(i)-EPSILON;
        
        numgrad(i) = (J(theta1)-J(theta2))/(2*EPSILON);
    end
        
    %% ---------------------------------------------------------------
    end
    

    如果用向量化计算,几十秒钟就运算出来了,最后结果如下:

  • 相关阅读:
    第十周上机作业
    第九周上机作业
    第八周作业
    第八周上机作业
    第七周作业
    第七周上机作业
    第六周作业
    第六周上机作业
    第五周上机作业
    第四周作业
  • 原文地址:https://www.cnblogs.com/boqun1991/p/4165689.html
Copyright © 2020-2023  润新知