void Evaluation::evalSplit (Trainer *trainer, ISamples &samples, ssi_real_t split) {
if (split <= 0 || split >= 1) {
ssi_err ("split must be a value between 0 and 1");
}
_trainer = trainer;
destroy_conf_mat ();
init_conf_mat (samples);
ssi_size_t *indices = new ssi_size_t[samples.getSize ()];
ssi_size_t *indices_count_lab = new ssi_size_t[samples.getClassSize ()];
ssi_size_t indices_count_all;
indices_count_all = 0;
for (ssi_size_t j = 0; j < samples.getClassSize (); j++) {
indices_count_lab[j] = 0;
}
ssi_size_t label;
ssi_size_t label_size;
for (ssi_size_t j = 0; j < samples.getSize (); j++) {
label = samples.get (j)->class_id;
label_size = samples.getSize (label);
if (++indices_count_lab[label] <= ssi_cast (ssi_size_t, label_size * split + 0.5f)) {
indices[indices_count_all++] = j;
}
}
SampleList strain;
SampleList stest;
// split off samples
ModelTools::SelectSampleList (samples, strain, stest, indices_count_all, indices);
_n_total = stest.getSize ();
_result_vec = new ssi_size_t[2*_n_total];
_result_vec_ptr = _result_vec;
// train with remaining samples
_trainer->release ();
if (_preproc_mode) {
_trainer->setPreprocMode (_preproc_mode, _n_streams_refs, _stream_refs);
} else if (_fselmethod) {
_trainer->setSelection (strain, _fselmethod, _pre_fselmethod, _n_pre_feature);
}
_trainer->train (strain);
// test with remaining samples
eval_h (stest);
delete[] indices;
delete[] indices_count_lab;
}
void Evaluation::evalKFold (Trainer *trainer, ISamples &samples, ssi_size_t k) {
// init confussion matrix
_trainer = trainer;
destroy_conf_mat ();
init_conf_mat (samples);
_n_total = samples.getSize ();
_result_vec = new ssi_size_t[2*_n_total];
_result_vec_ptr = _result_vec;
ssi_size_t *indices = new ssi_size_t[samples.getSize ()];
ssi_size_t *indices_count_lab = new ssi_size_t[samples.getClassSize ()];
ssi_size_t indices_count_all;
for (ssi_size_t i = 0; i < k; ++i) {
indices_count_all = 0;
for (ssi_size_t j = 0; j < samples.getClassSize (); j++) {
indices_count_lab[j] = 0;
}
ssi_size_t label;
for (ssi_size_t j = 0; j < samples.getSize (); j++) {
label = samples.get (j)->class_id;
if (++indices_count_lab[label] % k == i) {
indices[indices_count_all++] = j;
}
}
SampleList strain;
SampleList stest;
// split off i'th fold
ModelTools::SelectSampleList (samples, stest, strain, indices_count_all, indices);
// train with i'th fold
_trainer->release ();
if (_fselmethod) {
_trainer->setSelection (strain, _fselmethod, _pre_fselmethod, _n_pre_feature);
}
if (_preproc_mode) {
_trainer->setPreprocMode (_preproc_mode, _n_streams_refs, _stream_refs);
}
_trainer->train (strain);
// test with remaining samples
eval_h (stest);
}
delete[] indices;
delete[] indices_count_lab;
}
bool MyFusion::train (ssi_size_t n_models,
IModel **models,
ISamples &samples) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
ssi_size_t n_streams = samples.getStreamSize ();
if (n_streams != n_models) {
ssi_err ("#models (%u) differs from #streams (%u)", n_models, n_streams);
}
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) {
models[n_model]->train (samples, n_model);
}
}
_is_trained = true;
return true;
}
void AlgLibTools::Samples2MatrixWithClass (ISamples &samples,
ssi_size_t stream_id, ae_matrix* m) {
ae_int_t nfeatures = samples.get (0)->streams[stream_id]->dim;
ae_int_t nsamples = samples.getSize ();
ae_int_t i = 0;
ae_int_t j = 0;
ae_state state;
ae_matrix_clear(m);
ae_matrix_set_length(m, nsamples, nfeatures+1, &state);
ssi_sample_t *sample;
samples.reset ();
while (sample = samples.next ()) {
ssi_real_t *ptr = ssi_pcast (ssi_real_t, sample->streams[stream_id]->ptr);
for (j = 0; j <= nfeatures-1; j++)
{
m->ptr.pp_double[i][j] = ssi_cast (double, *ptr++);
}
m->ptr.pp_double[i][j] = ssi_cast (double, sample->class_id);
i++;
}
//delete sample;
}
void AlgLibTools::Samples2matrix (
ISamples &samples,
ssi_size_t stream_id,
ssi_size_t class_id,
ae_matrix* m,
ae_state *state)
{
ae_int_t nfeatures = samples.get (0)->streams[stream_id]->dim;
ae_int_t nsamples = samples.getSize (class_id);
ae_int_t i = 0;
ae_int_t j = 0;
ae_matrix_clear(m);
ae_matrix_set_length(m, nsamples, nfeatures, state);
ssi_sample_t *sample;
ISSelectClass samples_s (&samples);
samples_s.setSelection (class_id);
samples_s.reset ();
while (sample = samples_s.next ()) {
ssi_real_t *ptr = ssi_pcast (ssi_real_t, sample->streams[stream_id]->ptr);
for(j=0; j<=nfeatures-1; j++)
{
m->ptr.pp_double[i][j] = ssi_cast (double, *ptr++);
}
i++;
}
}
bool Fisher::build (ISamples &samples, ssi_size_t stream_index) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (isBuild ()) {
ssi_wrn ("already trained");
return false;
}
ae_state state;
ae_int_t info;
ae_matrix data;
ae_matrix_init (&data, 0, 0, DT_REAL, &state, ae_true);
// convert the samples to a matrix where the last column holds the class number to which the sample belongs
AlgLibTools::Samples2MatrixWithClass(samples, 0, &data);
_basis = new ae_matrix;
ae_matrix_init (_basis, 0, 0, DT_REAL, &state, ae_true);
fisherldan(&data,data.rows,data.cols-1 , samples.getClassSize(),&info,_basis,&state);
ae_matrix_clear (&data);
_is_build = true;
return true;
}
void Evaluation::evalLOO (Trainer *trainer, ISamples &samples) {
_trainer = trainer;
destroy_conf_mat ();
init_conf_mat (samples);
ssi_size_t n_samples = samples.getSize ();
_n_total = n_samples;
_result_vec = new ssi_size_t[2*_n_total];
_result_vec_ptr = _result_vec;
ssi_size_t itest = 0;
ssi_size_t *itrain = new ssi_size_t[n_samples - 1];
for (ssi_size_t nsample = 0; nsample < n_samples - 1; ++nsample) {
itrain[nsample] = nsample+1;
}
ISSelectSample strain (&samples);
ISSelectSample stest (&samples);
strain.setSelection (n_samples-1, itrain);
stest.setSelection (1, &itest);
_trainer->release ();
if (_fselmethod) {
_trainer->setSelection (strain, _fselmethod, _pre_fselmethod, _n_pre_feature);
}
if (_preproc_mode) {
_trainer->setPreprocMode (_preproc_mode, _n_streams_refs, _stream_refs);
}
_trainer->train (strain);
eval_h (stest);
for (ssi_size_t nsample = 1; nsample < n_samples; ++nsample) {
itrain[nsample-1] = nsample-1;
itest = nsample;
strain.setSelection (n_samples-1, itrain);
stest.setSelection (1, &itest);
_trainer->release ();
if (_fselmethod) {
_trainer->setSelection (strain, _fselmethod, _pre_fselmethod, _n_pre_feature);
}
if (_preproc_mode) {
_trainer->setPreprocMode (_preproc_mode, _n_streams_refs, _stream_refs);
}
_trainer->train (strain);
eval_h (stest);
}
delete [] itrain;
}
bool MyModel::train (ISamples &samples,
ssi_size_t stream_index) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
_n_classes = samples.getClassSize ();
_n_features = samples.getStream (stream_index).dim;
_centers = new ssi_real_t *[_n_classes];
for (ssi_size_t i = 0; i < _n_classes; i++) {
_centers[i] = new ssi_real_t[_n_features];
for (ssi_size_t j = 0; j < _n_features; j++) {
_centers[i][j] = 0;
}
}
ssi_sample_t *sample;
samples.reset ();
ssi_real_t *ptr = 0;
while (sample = samples.next ()) {
ssi_size_t id = sample->class_id;
ptr = ssi_pcast (ssi_real_t, sample->streams[stream_index]->ptr);
for (ssi_size_t j = 0; j < _n_features; j++) {
_centers[id][j] += ptr[j];
}
}
for (ssi_size_t i = 0; i < _n_classes; i++) {
ssi_size_t num = samples.getSize (i);
for (ssi_size_t j = 0; j < _n_features; j++) {
_centers[i][j] /= num;
}
}
return true;
}
bool SimpleKNN::train (ISamples &samples,
ssi_size_t stream_index) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (samples.getSize () < _options.k) {
ssi_wrn ("sample list has less than '%u' entries", _options.k);
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
_n_classes = samples.getClassSize ();
_n_samples = samples.getSize ();
_n_features = samples.getStream (stream_index).dim;
_data = new ssi_real_t[_n_features*_n_samples];
_classes = new ssi_size_t[_n_samples];
ssi_sample_t *sample;
samples.reset ();
ssi_real_t *data_ptr = _data;
ssi_size_t *class_ptr = _classes;
ssi_stream_t *stream_ptr = 0;
ssi_size_t bytes_to_copy = _n_features * sizeof (ssi_real_t);
while (sample = samples.next ()) {
memcpy (data_ptr, sample->streams[stream_index]->ptr, bytes_to_copy);
*class_ptr++ = sample->class_id;
data_ptr += _n_features;
}
return true;
}
void Evaluation::eval (Trainer *trainer, ISamples &samples) {
// init confussion matrix
_trainer = trainer;
destroy_conf_mat ();
init_conf_mat (samples);
_n_total = samples.getSize ();
_result_vec = new ssi_size_t[2*_n_total];
_result_vec_ptr = _result_vec;
// call helper function
eval_h (samples);
}
void Evaluation::eval (IFusion &fusion, ssi_size_t n_models, IModel **models, ISamples &samples) {
// init confussion matrix
_trainer = 0;
destroy_conf_mat ();
init_conf_mat (samples);
ssi_size_t n_classes = samples.getClassSize ();
ssi_real_t *probs = new ssi_real_t[n_classes];
_n_total = samples.getSize ();
_result_vec = new ssi_size_t[2*_n_total];
_result_vec_ptr = _result_vec;
samples.reset ();
const ssi_sample_t *sample = 0;
while (sample = samples.next ()) {
ssi_size_t real_index = sample->class_id;
*_result_vec_ptr++ = real_index;
if (fusion.forward (n_models, models, sample->num, sample->streams, n_classes, probs)) {
ssi_size_t max_ind = 0;
ssi_real_t max_val = probs[0];
for (ssi_size_t i = 1; i < n_classes; i++) {
if (probs[i] > max_val) {
max_val = probs[i];
max_ind = i;
}
}
*_result_vec_ptr++ = max_ind;
_conf_mat_ptr[real_index][max_ind]++;
_n_classified++;
} else if (!_allow_unclassified) {
ssi_size_t max_ind = _default_class_id;
*_result_vec_ptr++ = max_ind;
_conf_mat_ptr[real_index][max_ind]++;
_n_classified++;
} else {
*_result_vec_ptr++ = SSI_ISAMPLES_GARBAGE_CLASS_ID;
_n_unclassified++;
}
}
delete[] probs;
}
bool MajorityVoting::train (ssi_size_t n_models,
IModel **models,
ISamples &samples) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (samples.getStreamSize () != n_models) {
ssi_wrn ("#models (%u) differs from #streams (%u)", n_models, samples.getStreamSize ());
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
_n_streams = samples.getStreamSize ();
_n_classes = samples.getClassSize ();
_n_models = n_models;
if (samples.hasMissingData ()) {
ISMissingData samples_h (&samples);
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) {
samples_h.setStream (n_model);
models[n_model]->train (samples_h, n_model);
}
}
}
else{
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) { models[n_model]->train (samples, n_model); }
}
}
return true;
}
bool SimpleFusion::train (ssi_size_t n_models,
IModel **models,
ISamples &samples) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
ssi_size_t n_streams = samples.getStreamSize ();
if (n_streams != 1 && n_streams != n_models) {
ssi_err ("#models (%u) differs from #streams (%u)", n_models, n_streams);
}
if (samples.hasMissingData ()) {
ISMissingData samples_h (&samples);
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) {
samples_h.setStream(n_streams == 1 ? 0 : n_model);
models[n_model]->train (samples_h, n_model);
}
}
} else {
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) {
models[n_model]->train(samples, n_streams == 1 ? 0 : n_model);
}
}
}
_is_trained = true;
return true;
}
bool FeatureFusion::train (ssi_size_t n_models,
IModel **models,
ISamples &samples) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
_n_streams = samples.getStreamSize ();
_n_classes = samples.getClassSize ();
_n_models = n_models;
//initialize weights
ssi_real_t **weights = new ssi_real_t*[n_models];
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
weights[n_model] = new ssi_real_t[_n_classes+1];
}
if (samples.hasMissingData ()) {
_handle_md = true;
ISMissingData samples_h (&samples);
Evaluation eval;
if (ssi_log_level >= SSI_LOG_LEVEL_DEBUG) {
ssi_print("\nMissing data detected.\n");
}
//models[0] is featfuse_model, followed by singlechannel_models
ISMergeDim ffusionSamples (&samples);
ISMissingData ffusionSamples_h (&ffusionSamples);
ffusionSamples_h.setStream(0);
if (!models[0]->isTrained ()) { models[0]->train (ffusionSamples_h, 0); }
if (ssi_log_level >= SSI_LOG_LEVEL_DEBUG) {
eval.eval (*models[0], ffusionSamples_h, 0);
eval.print();
}
//dummy weights for fused model
for (ssi_size_t n_class = 0; n_class < _n_classes; n_class++) {
weights[0][n_class] = 0.0f;
}
weights[0][_n_classes] = 0.0f;
for (ssi_size_t n_model = 1; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) {
samples_h.setStream (n_model - 1);
models[n_model]->train (samples_h, n_model - 1);
}
eval.eval (*models[n_model], samples_h, n_model - 1);
if (ssi_log_level >= SSI_LOG_LEVEL_DEBUG) {
eval.print();
}
for (ssi_size_t n_class = 0; n_class < _n_classes; n_class++) {
weights[n_model][n_class] = eval.get_class_prob (n_class);
}
weights[n_model][_n_classes] = eval.get_classwise_prob ();
}
//calculate fillers
_filler = new ssi_size_t[_n_streams];
for (ssi_size_t n_fill = 0; n_fill < _n_streams; n_fill++) {
_filler[n_fill] = 1;
ssi_real_t filler_weight = weights[1][_n_classes];
for (ssi_size_t n_model = 2; n_model < n_models; n_model++) {
if (filler_weight < weights[n_model][_n_classes]) {
_filler[n_fill] = n_model;
filler_weight = weights[n_model][_n_classes];
}
}
weights[_filler[n_fill]][_n_classes] = 0.0f;
}
if (ssi_log_level >= SSI_LOG_LEVEL_DEBUG) {
ssi_print("\nfiller:\n");
for (ssi_size_t n_model = 0; n_model < _n_streams; n_model++) {
ssi_print("%d ", _filler[n_model]);
}ssi_print("\n");
}
}
else{
_handle_md = false;
if (ssi_log_level >= SSI_LOG_LEVEL_DEBUG) {
ssi_print("\nNo missing data detected.\n");
}
ISMergeDim ffusionSamples (&samples);
if (!models[0]->isTrained ()) { models[0]->train (ffusionSamples, 0); }
//dummy
_filler = new ssi_size_t[_n_streams];
for (ssi_size_t n_fill = 0; n_fill < _n_streams; n_fill++) {
_filler[n_fill] = 0;
}
}
if (weights) {
for (ssi_size_t n_model = 0; n_model < _n_models; n_model++) {
delete[] weights[n_model];
}
delete[] weights;
weights = 0;
}
return true;
}
bool WeightedMajorityVoting::train (ssi_size_t n_models,
IModel **models,
ISamples &samples) {
if (samples.getSize () == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (samples.getStreamSize () != n_models) {
ssi_wrn ("#models (%u) differs from #streams (%u)", n_models, samples.getStreamSize ());
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
_n_streams = samples.getStreamSize ();
_n_classes = samples.getClassSize ();
_n_models = n_models;
_weights = new ssi_real_t*[n_models];
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
_weights[n_model] = new ssi_real_t[_n_classes+1];
}
if (samples.hasMissingData ()) {
ISMissingData samples_h (&samples);
Evaluation eval;
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) {
samples_h.setStream (n_model);
models[n_model]->train (samples_h, n_model);
}
eval.eval (*models[n_model], samples_h, n_model);
for (ssi_size_t n_class = 0; n_class < _n_classes; n_class++) {
_weights[n_model][n_class] = eval.get_class_prob (n_class);
}
_weights[n_model][_n_classes] = eval.get_classwise_prob ();
}
}
else{
Evaluation eval;
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
if (!models[n_model]->isTrained ()) { models[n_model]->train (samples, n_model); }
eval.eval (*models[n_model], samples, n_model);
for (ssi_size_t n_class = 0; n_class < _n_classes; n_class++) {
_weights[n_model][n_class] = eval.get_class_prob (n_class);
}
_weights[n_model][_n_classes] = eval.get_classwise_prob ();
}
}
if (ssi_log_level >= SSI_LOG_LEVEL_DEBUG) {
ssi_print("\nClassifier Weights: \n");
for (ssi_size_t n_model = 0; n_model < n_models; n_model++) {
for (ssi_size_t n_class = 0; n_class < _n_classes; n_class++) {
ssi_print ("%f ", _weights[n_model][n_class]);
}
ssi_print ("%f\n", _weights[n_model][_n_classes]);
}
}
return true;
}
bool SVM::train (ISamples &samples,
ssi_size_t stream_index) {
if (_options.seed > 0) {
srand(_options.seed);
} else {
srand(ssi_time_ms());
}
ISamples *s_balance = 0;
switch (_options.balance) {
case BALANCE::OFF: {
s_balance = &samples;
break;
}
case BALANCE::OVER: {
s_balance = new ISOverSample(&samples);
ssi_pcast(ISOverSample, s_balance)->setOver(ISOverSample::RANDOM);
ssi_msg(SSI_LOG_LEVEL_BASIC, "balance training set '%u' -> '%u'", samples.getSize(), s_balance->getSize());
break;
}
case BALANCE::UNDER: {
s_balance = new ISUnderSample(&samples);
ssi_pcast(ISUnderSample, s_balance)->setUnder(ISUnderSample::RANDOM);
ssi_msg(SSI_LOG_LEVEL_BASIC, "balance training set '%u' -> '%u'", samples.getSize(), s_balance->getSize());
break;
}
}
_n_samples = s_balance->getSize();
if (_n_samples == 0) {
ssi_wrn ("empty sample list");
return false;
}
if (isTrained ()) {
ssi_wrn ("already trained");
return false;
}
_n_classes = s_balance->getClassSize();
_n_features = s_balance->getStream(stream_index).dim;
ssi_size_t elements = _n_samples * (_n_features + 1);
init_class_names(*s_balance);
_problem = new svm_problem;
_problem->l = ssi_cast (int, _n_samples);
_problem->y = new double[_problem->l];
_problem->x = new svm_node *[_problem->l];
s_balance->reset();
ssi_sample_t *sample;
int n_sample = 0;
float *ptr = 0;
svm_node *node = 0;
while (sample = s_balance->next()) {
ptr = ssi_pcast (float, sample->streams[stream_index]->ptr);
_problem->x[n_sample] = new svm_node[_n_features + 1];
_problem->y[n_sample] = ssi_cast (float, sample->class_id);
node = _problem->x[n_sample];
for (ssi_size_t nfeat = 0; nfeat < _n_features; nfeat++) {
node->index = nfeat+1;
node->value = *ptr;
ptr++;
++node;
}
node->index = -1;
++n_sample;
}
if(_options.params.gamma == 0 && _n_features > 0) {
_options.params.gamma = 1.0 / _n_features;
}
if (_options.params.kernel_type == PRECOMPUTED) {
int max_index = ssi_cast (int, _n_features);
for (int i = 0; i < _problem->l; i++) {
if (_problem->x[i][0].index != 0) {
ssi_err ("wrong input format: first column must be 0:sample_serial_number");
}
if ((int)_problem->x[i][0].value <= 0 || (int)_problem->x[i][0].value > max_index) {
ssi_err ("wrong input format: sample_serial_number out of range");
}
}
}