function [model] = oaasmo(data,labels,ker,arg,C,epsilon,tolerance,info)% OAASMO One-Against-All multi-class decomposition solved by SMO.% [model] = oaasmo(X,I,ker,arg,C,epsilon,tolerance,info)%% OAASMO solves multi-class Support Vector Problem problem %  by the use of Sequential Minimal Optimizer.%  The input K-class problem is decomposed onto %  K binary problems: i-th class against remaining (K-1)%  classes. The binary problems are solved by the use %  of the SMO which results to the K decision rules.%  % Mandatory inputs:%  data [NxL] contains L training N-dimensional patterns.%  labels [1xL] labels of the patterns; integers from 1 to K.%  ker [string] kernel identifier (see "help smoc").%  arg [...] arguments of given kernel (see "help smoc").%  C [1x1] trade-off between maximal margin and training error.%% Optional inputs:%  epsilon [real] tolerance of KKT-conditions fulfilment %    (default is 0.001).%  tolerance [real] minimal change of optimized Lagrangeians %    (default is 0.001).%  info [int] if info is 1 then prints which decision rule is%    being built (default is 0).%% Output: %  model [struct] contains found multi-class SVM classifier. %  % For instance malpha{i} and mbias{i} are the multipliers and% bias for i-th binary problem: i-th class against the others.%% See also OAACLASS, SMO, SVM.%% Statistical Pattern Recognition Toolbox, Vojtech Franc, Vaclav Hlavac% (c) Czech Technical University Prague, http://cmp.felk.cvut.cz% Modifications%  2-Sep-2002, VF%  21-Oct-2001, V.Franc%  18-September-2001, V. Franc, created% -- Processing of the input arguments ----------------if nargin < 5,   error('Not enough input arguments.');endif nargin < 6,   epsilon = 0.001;endif nargin < 7,   tolerance = 0.001;endif nargin < 8,   info = 0;end%---------------------------------% get dimensions[dim,num_data]=size(data);num_classes=max(labels);%---------------------------------model.name = 'One-Against-All SVM classifier';model.num_classes = num_classes;model.num_rules = num_classes;model.rule = cell(model.num_rules);model.kercnt = 0;model.btime = 0;tmp_alpha=zeros(1,num_data);for i=1:num_classes,   if info==1,     fprintf(1,'Building rule %d of %d\n', i,num_classes);   end     dich_labels=multi2dicho( labels, i );      t=cputime;      [Alpha,bias,nsv,kercnt]=...       smo(data,dich_labels,ker,arg,C,epsilon,tolerance);      model.btime= model.btime + cputime-t;   model.kercnt = model.kercnt + kercnt;   model.rule{i}.Alpha = Alpha;   model.rule{i}.bias = bias;      tmp_alpha = tmp_alpha+Alpha;endmodel.SVM.C = C;model.SVM.kernel = ker;model.SVM.arg = arg;model.trn_data = data;model.trn_labels = labels;model.kercnt = kercnt;%model.trnerr = ;model.nsv = length(find(tmp_alpha));  return;