RBM代码实现MATLAB

This is a small library that can train Restricted Boltzmann Machines, and also Deep Belief Networks of stacked RBM's. Train RBM's: %train an RBM with binary visible units and 500 binary hidden model= rbmBB(data, 500); %visualize the learned weights visualize(model.W); Do classification: model= rbmFit(data, 500, labels); prediction= rbmPredict(model, testdata); Train a Deep Belief Network with 500,500,2000 architecture for classification: models= dbnFit(data, [500 500 2000], labels); prediction= dbnPredict(models, testdata); see included example code for more I can be contacted on andrej.karpathy@ gmail. NOTE: This was a class project that I worked on for 1 month and then abandoned development for almost 4 years ago. Please do not send me specific questions about issues with the code or questions on how to do something. I only put this code online in hope that it can be useful to others but cannot fully support it. If you would like pointers to more actively maintained implementations, have a look here (https://github.com/rasmusbergpalm/DeepLearnToolbox) or maybe here (https://github.com/lisa-lab/DeepLearningTutorials) Sorry and best of luck! 原文：http://code.google.com/p/matrbm/

load mnist_classify;

%% Train RBM for classification
%train rbm with 100 hidden units
m=rbmFit（data100labels‘verbose‘true）;
yhat=rbmPredict（mtestdata）;

%print error
fprintf（‘Classification error using RBM with 100 hiddens is %f\n‘ ...
    sum（yhat~=testlabels）/length（yhat））;

%visualize weights
figure（1）
visualize（m.W）;
title（‘learned weights‘）;

%visualize the mislabeled cases. Note the transpose. Visualize assumes DxN
%as is the case for weights
figure（2）
visualize（data（yhat~=testlabels:）‘）;
title（‘classification mistakes for RBM with 100 hiddens‘）;
drawnow;

%% Train model and denoise images
m2= rbmBB（data100‘verbose‘true）;

%distort 100 images around by setting 95% to random noise
imgs=testdata（1:100:）;
b=rand（size（imgs））>0.95;
noised=imgs;
r=rand（size（imgs））;
noised（b）=r（b）;

%reconstruct the images by going up down then up again using learned model
up = rbmVtoH（m2 noised）;
down= rbmHtoV（m2 up）;

%figure
z1=visualize（noised‘）;
z2=visualize（down‘）;

figure（3）
imshow（[z1 z2]）
title（‘denoising 95% noise with RBM with 100 hidden units‘）;
drawnow;
%% Train a DBN
op.verbose=true;
models=dbnFit（data[100 100]labelsopop）;
yhat2=dbnPredict（modelstestdata）;

%print error
fprintf（‘Classification error using DBN with 100-100 hiddens is %f\n‘ ...
    sum（yhat2~=testlabels）/length（yhat2））;

%visualize weights
figure（4）
subplot（121）
visualize（models{1}.W）;
title（‘learned weights on DBN layer 1‘）;
subplot（122）
visualize（models{2}.W）;
title（‘learned weights on DBN layer 2‘）;

%visualize the mislabeled cases. Note the transpose. Visualize assumes DxN
%as is the case for weights
figure（5）
visualize（data（yhat2~=testlabels:）‘）;
title（‘classification mistakes for DBN with 100-100 hiddens‘）;

 属性            大小     日期    时间   名称
----------- ---------  ---------- -----  ----

     文件       1796  2010-10-31 13:00  RBMLIB\examplecode.m

     文件    2974337  2010-10-31 13:00  RBMLIB\mnist_classify.mat

     文件       1577  2010-10-31 13:01  RBMLIB\RBM\dbnFit.m

     文件        495  2010-10-31 13:01  RBMLIB\RBM\dbnPredict.m

     文件        409  2010-10-31 13:01  RBMLIB\RBM\interweave.m

     文件         65  2010-10-31 13:01  RBMLIB\RBM\logistic.m

     文件        977  2010-10-31 13:01  RBMLIB\RBM\nunique.m

     文件        690  2010-10-31 13:01  RBMLIB\RBM\prepareArgs.m

     文件       3819  2010-10-31 13:01  RBMLIB\RBM\process_options.m

     文件       5217  2010-10-31 13:01  RBMLIB\RBM\rbmBB.m

     文件       6147  2010-10-31 13:16  RBMLIB\RBM\rbmFit.m

     文件        358  2010-10-31 13:01  RBMLIB\RBM\rbmHtoV.m

     文件        877  2010-10-31 13:22  RBMLIB\RBM\rbmPredict.m

     文件        355  2010-10-31 13:01  RBMLIB\RBM\rbmVtoH.m

     文件        286  2010-10-31 13:01  RBMLIB\RBM\softmaxPmtk.m

     文件        371  2010-10-31 13:01  RBMLIB\RBM\softmax_sample.m

     文件        750  2010-10-31 13:01  RBMLIB\RBM\visualize.m

     文件      13554  2015-03-23 21:12  RBMLIB\readme.docx

     目录          0  2015-03-23 21:10  RBMLIB\RBM

     目录          0  2015-03-23 21:12  RBMLIB

----------- ---------  ---------- -----  ----

              3012080                    20