Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

add mushroom classification

Browse Source
This commit is contained in:
tqchen 2014-02-24 22:25:43 -08:00
parent 4aa4faa625
commit 9d6ef11eb5
2 changed files with 26 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
demo/mushroom/mushroom.conf
View File

@ -1,4 +1,4 @@
num_round=10
num_round=2
save_period=0
@ -6,7 +6,7 @@ data = "agaricus.txt.train"
eval[test] = "agaricus.txt.test"
booster_type = 0
loss_type = 1
loss_type = 2
bst:num_feature=126
bst:eta=1.0

25
regression/xgboost_reg.h
View File

@ -172,6 +172,7 @@ namespace xgboost{
enum LossType{
kLinearSquare = 0,
kLogisticNeglik = 1,
kLogisticClassify = 2
};
/*! \brief training parameter for regression */
@ -211,6 +212,7 @@ namespace xgboost{
inline float PredTransform( float x ){
switch( loss_type ){
case kLinearSquare: return x;
case kLogisticClassify:
case kLogisticNeglik: return 1.0f/(1.0f + expf(-x));
default: utils::Error("unknown loss_type"); return 0.0f;
}
@ -225,7 +227,8 @@ namespace xgboost{
inline float FirstOrderGradient( float predt, float label ) const{
switch( loss_type ){
case kLinearSquare: return predt - label;
case 1: return predt - label;
case kLogisticClassify:
case kLogisticNeglik: return predt - label;
default: utils::Error("unknown loss_type"); return 0.0f;
}
}
@ -238,6 +241,7 @@ namespace xgboost{
inline float SecondOrderGradient( float predt, float label ) const{
switch( loss_type ){
case kLinearSquare: return 1.0f;
case kLogisticClassify:
case kLogisticNeglik: return predt * ( 1 - predt );
default: utils::Error("unknown loss_type"); return 0.0f;
}
@ -253,6 +257,7 @@ namespace xgboost{
switch( loss_type ){
case kLinearSquare: return SquareLoss(preds,labels);
case kLogisticNeglik: return NegLoglikelihoodLoss(preds,labels);
case kLogisticClassify: return ClassificationError(preds, labels);
default: utils::Error("unknown loss_type"); return 0.0f;
}
}
@ -284,6 +289,24 @@ namespace xgboost{
ans -= labels[i] * logf(preds[i]) + ( 1 - labels[i] ) * logf(1 - preds[i]);
return ans;
}
/*!
* \brief calculating the ClassificationError loss, given the predictions and labels
* \param preds the given predictions
* \param labels the given labels
* \return the summation of square loss
*/
inline float ClassificationError(const std::vector<float> &preds, const std::vector<float> &labels) const{
int nerr = 0;
for(size_t i = 0; i < preds.size(); i++){
if( preds[i] > 0.5f ){
if( labels[i] < 0.5f ) nerr ++;
}else{
if( labels[i] > 0.5f ) nerr ++;
}
}
return (float)nerr/preds.size();
}
};
private:
int silent;