ML之xgboost：基于xgboost(5f-CrVa)算法对HiggsBoson数据集(Kaggle竞赛)训练实现二分类预测(基于训练好的模型进行新数据预测)

输出结果

设计思路

核心代码

xgmat = xgb.DMatrix( data, missing = -999.0 )  

bst = xgb.Booster({'nthread':8}, model_file = modelfile)

res  = [ ( int(idx[i]), ypred[i] ) for i in range(len(ypred)) ]  

rorder = {}

for k, v in sorted( res, key = lambda x:-x[1] ):                

   rorder[ k ] = len(rorder) + 1

# write out predictions

ntop = int( threshold_ratio * len(rorder ) )

fo = open(outfile, 'w')                              

nhit = 0

ntot = 0

fo.write('EventId,RankOrder,Class\n')                

for k, v in res:                                    

   if rorder[k] <= ntop:

       lb = 's'

       nhit += 1

   else:

       lb = 'b'

   # change output rank order to follow Kaggle convention

   fo.write('%s,%d,%s\n' % ( k,  len(rorder)+1-rorder[k], lb ) )

   ntot += 1

fo.close()