view reproduce_AES53rd/rerun_svm_table3/svm_table3.m @ 0:e9a9cd732c1e tip

first hg version after svn

% ---
% this script trains similarity measures using RBM and SVM as in Table 3
% please note that the RBM training is a probabilistic process.
% Here, training is done on 20 random initialisations of RBM features ,
% the test results corresponding to the RBM with the best training result are then
% returned.
% ---

% ---
% vary feature parameters of mixed features
% ---

global globalvars;
globalvars.debug = 3;
        
ftype = 'MTTMixedFeatureSonRBM'; %'MTTMixedFeatureStober11Genre';

fparams_all = struct(...
            ... % ---
            ... % these are SONfeatRaw parameters
            ... % ---
            'son_filename',{{'rel_music_raw_features+simdata_ISMIR12.mat'}}, ...
            'son_conf', 1:5, ...
             ... % --- 
            ... % Following: RBM params
            ... % ---
            'norm_pre_rbm', 0, ... % norm before RBM?
            'norm_post_rbm',0, ... % norm before RBM?
            'rbm_hidNum',[1000], ...   % number of hidden units % 500
            'rbm_eNum', 100, ...
            'rbm_bNum', 1, ...
            'rbm_gNum', 1, ...
            'rbm_lrate1' , [0.05], ... % initial learning rate % 0.01
            'rbm_lrate2', [0.10], ... %  learning rate, %0.05  
            'rbm_momentum', [0.1], ... % 0.5
            'rbm_cost', [0.00002], ... % cost function
            'rbm_N', 50, ...
            'rbm_MAX_INC', 10 ...
        );

% ---
% vary parameters for svmlight
% ---    

trainparams_all = struct(...
            'C', [1], ...%    
            'weighted', [0], ...
            'dataset', {{'comp_partBinData_ISMIR12_01.mat'}}, ...
            'inctrain', 0 ...
            ... % this optional
            ... %'deltafun', {{'conv_subspace_delta'}}, ...
            ... %'deltafun_params', {{{[1],[0]},{[5],[1]},{[10],[1]},{[20],[1]},{[30],[1]},{[50],[1]},{[70],[1]}}} ... % normalisation improves results
            );

% set training function
trainfun = @svmlight_wrapper;


% create test directory
akt_dir = migrate_to_test_dir();


% call eval
out = test_generic_features_parameters_crossval...
    (fparams_all, trainparams_all, trainfun, ftype);

% ---
% check training results and select best RBM according to trainign data
% ---
svm_train_performances = [out(:).mean_ok_train];
[bestTrain, idx] = max(svm_train_performances(1,:));
result = out(idx);

% get corresponding test performance
svm_test_performance = result.mean_ok_test(1,:);
fprintf('SVM RBM Test/Train Result=%f / %f\n',svm_test_performance*100,bestTrain*100);