LibSVMMultiClass.cpp

Go to the documentation of this file.
00001 /*
00002  * This program is free software; you can redistribute it and/or modify
00003  * it under the terms of the GNU General Public License as published by
00004  * the Free Software Foundation; either version 3 of the License, or
00005  * (at your option) any later version.
00006  *
00007  * Written (W) 1999-2009 Soeren Sonnenburg
00008  * Copyright (C) 1999-2009 Fraunhofer Institute FIRST and Max-Planck-Society
00009  */
00010 
00011 #include "classifier/svm/LibSVMMultiClass.h"
00012 #include "lib/io.h"
00013 
00014 using namespace shogun;
00015 
00016 CLibSVMMultiClass::CLibSVMMultiClass(LIBSVM_SOLVER_TYPE st)
00017 : CMultiClassSVM(ONE_VS_ONE), model(NULL), solver_type(st)
00018 {
00019 }
00020 
00021 CLibSVMMultiClass::CLibSVMMultiClass(float64_t C, CKernel* k, CLabels* lab)
00022 : CMultiClassSVM(ONE_VS_ONE, C, k, lab), model(NULL), solver_type(LIBSVM_C_SVC)
00023 {
00024 }
00025 
00026 CLibSVMMultiClass::~CLibSVMMultiClass()
00027 {
00028     //SG_PRINT("deleting LibSVM\n");
00029 }
00030 
00031 bool CLibSVMMultiClass::train(CFeatures* data)
00032 {
00033     struct svm_node* x_space;
00034 
00035     problem = svm_problem();
00036 
00037     ASSERT(labels && labels->get_num_labels());
00038     int32_t num_classes = labels->get_num_classes();
00039     problem.l=labels->get_num_labels();
00040     SG_INFO( "%d trainlabels, %d classes\n", problem.l, num_classes);
00041 
00042     if (data)
00043     {
00044         if (labels->get_num_labels() != data->get_num_vectors())
00045             SG_ERROR("Number of training vectors does not match number of labels\n");
00046         kernel->init(data, data);
00047     }
00048 
00049     problem.y=new float64_t[problem.l];
00050     problem.x=new struct svm_node*[problem.l];
00051     problem.pv=new float64_t[problem.l];
00052     problem.C=new float64_t[problem.l];
00053 
00054     x_space=new struct svm_node[2*problem.l];
00055 
00056     for (int32_t i=0; i<problem.l; i++)
00057     {
00058         problem.pv[i]=-1.0;
00059         problem.y[i]=labels->get_label(i);
00060         problem.x[i]=&x_space[2*i];
00061         x_space[2*i].index=i;
00062         x_space[2*i+1].index=-1;
00063     }
00064 
00065     ASSERT(kernel);
00066 
00067     param.svm_type=solver_type; // C SVM or NU_SVM
00068     param.kernel_type = LINEAR;
00069     param.degree = 3;
00070     param.gamma = 0;    // 1/k
00071     param.coef0 = 0;
00072     param.nu = get_nu(); // Nu
00073     param.kernel=kernel;
00074     param.cache_size = kernel->get_cache_size();
00075     param.C = get_C1();
00076     param.eps = epsilon;
00077     param.p = 0.1;
00078     param.shrinking = 1;
00079     param.nr_weight = 0;
00080     param.weight_label = NULL;
00081     param.weight = NULL;
00082     param.use_bias = get_bias_enabled();
00083 
00084     const char* error_msg = svm_check_parameter(&problem,&param);
00085 
00086     if(error_msg)
00087         SG_ERROR("Error: %s\n",error_msg);
00088 
00089     model = svm_train(&problem, &param);
00090 
00091     if (model)
00092     {
00093         ASSERT(model->nr_class==num_classes);
00094         ASSERT((model->l==0) || (model->l>0 && model->SV && model->sv_coef));
00095         create_multiclass_svm(num_classes);
00096 
00097         int32_t* offsets=new int32_t[num_classes];
00098         offsets[0]=0;
00099 
00100         for (int32_t i=1; i<num_classes; i++)
00101             offsets[i] = offsets[i-1]+model->nSV[i-1];
00102 
00103         int32_t s=0;
00104         for (int32_t i=0; i<num_classes; i++)
00105         {
00106             for (int32_t j=i+1; j<num_classes; j++)
00107             {
00108                 int32_t k, l;
00109 
00110                 float64_t sgn=1;
00111                 if (model->label[i]>model->label[j])
00112                     sgn=-1;
00113 
00114                 int32_t num_sv=model->nSV[i]+model->nSV[j];
00115                 float64_t bias=-model->rho[s];
00116 
00117                 ASSERT(num_sv>0);
00118                 ASSERT(model->sv_coef[i] && model->sv_coef[j-1]);
00119 
00120                 CSVM* svm=new CSVM(num_sv);
00121 
00122                 svm->set_bias(sgn*bias);
00123 
00124                 int32_t sv_idx=0;
00125                 for (k=0; k<model->nSV[i]; k++)
00126                 {
00127                     svm->set_support_vector(sv_idx, model->SV[offsets[i]+k]->index);
00128                     svm->set_alpha(sv_idx, sgn*model->sv_coef[j-1][offsets[i]+k]);
00129                     sv_idx++;
00130                 }
00131 
00132                 for (k=0; k<model->nSV[j]; k++)
00133                 {
00134                     svm->set_support_vector(sv_idx, model->SV[offsets[j]+k]->index);
00135                     svm->set_alpha(sv_idx, sgn*model->sv_coef[i][offsets[j]+k]);
00136                     sv_idx++;
00137                 }
00138 
00139                 int32_t idx=0;
00140 
00141                 if (sgn>0)
00142                 {
00143                     for (k=0; k<model->label[i]; k++)
00144                         idx+=num_classes-k-1;
00145 
00146                     for (l=model->label[i]+1; l<model->label[j]; l++)
00147                         idx++;
00148                 }
00149                 else
00150                 {
00151                     for (k=0; k<model->label[j]; k++)
00152                         idx+=num_classes-k-1;
00153 
00154                     for (l=model->label[j]+1; l<model->label[i]; l++)
00155                         idx++;
00156                 }
00157 
00158 
00159 //              if (sgn>0)
00160 //                  idx=((num_classes-1)*model->label[i]+model->label[j])/2;
00161 //              else
00162 //                  idx=((num_classes-1)*model->label[j]+model->label[i])/2;
00163 //
00164                 SG_DEBUG("svm[%d] has %d sv (total: %d), b=%f label:(%d,%d) -> svm[%d]\n", s, num_sv, model->l, bias, model->label[i], model->label[j], idx);
00165 
00166                 set_svm(idx, svm);
00167                 s++;
00168             }
00169         }
00170 
00171         CSVM::set_objective(model->objective);
00172 
00173         delete[] offsets;
00174         delete[] problem.x;
00175         delete[] problem.y;
00176         delete[] x_space;
00177 
00178         svm_destroy_model(model);
00179         model=NULL;
00180 
00181         return true;
00182     }
00183     else
00184         return false;
00185 }
00186 

SHOGUN Machine Learning Toolbox - Documentation