training data 0.00263262
negative training 0.002332536
training set 0.0023282479999999998
negative data 0.002157796
positive training 0.002038997
data different 0.002038449
training examples 0.002034511
learning problem 0.001968028
training documents 0.001963604
same data 0.0018884890000000001
training sample 0.00186732
positive data 0.001864257
negative set 0.001853424
test data 0.001837382
negative class 0.001824726
training instances 0.001796543
learning method 0.0017961879999999998
training samples 0.001773568
learning model 0.001763993
learning algorithm 0.0017583339999999999
text classification 0.001757945
learning methods 0.0017336229999999999
original training 0.001725081
tive training 0.001717772
ing data 0.001715083
weighted training 0.001690484
training distribu 0.001690203
text classifier 0.0016701340000000002
supervised learning 0.0016655849999999998
ample training 0.001660761
data sets 0.001645197
data figure 0.001637345
classification results 0.0015838879999999999
base learning 0.00157264
learning algorithms 0.0015680639999999999
machine learning 0.0015664799999999999
traditional learning 0.0015646759999999999
positive set 0.001559885
negative examples 0.0015596870000000001
unlabeled set 0.0015577249999999998
learning technique 0.001555784
various data 0.001550811
data distributions 0.001550606
learning techniques 0.001541597
negative test 0.001537298
test set 0.00153301
positive class 0.001531187
learning models 0.001530612
noisy data 0.0015154539999999998
data points 0.001509247
newsgroup data 0.001494339
transfer learning 0.001493341
reuters data 0.001492993
fer learning 0.001491577
negative documents 0.0014887799999999999
benchmark data 0.0014886909999999999
reuter data 0.001476318
incomplete data 0.001475377
different feature 0.001466971
word feature 0.0014588959999999999
class documents 0.0014557939999999998
class label 0.0014444850000000001
classification algorithm 0.0014441340000000001
classification methods 0.001419423
set methods 0.001412991
training 0.00140368
first classifier 0.001386751
problem work 0.001377546
class labels 0.0013747899999999999
classification system 0.00137022
class distribution 0.001364703
negative classes 0.001360854
negative document 0.001359172
final classification 0.001327715
bayesian classification 0.00132708
negative instances 0.001321719
good set 0.001317984
different case 0.001294555
classifier parameters 0.001288017
svm classifier 0.0012879
different distribution 0.001278342
practical classification 0.0012679359999999999
positive examples 0.001266148
unlabeled examples 0.001263988
classification problems 0.001260656
experiment set 0.001260147
traditional classification 0.001250476
learning 0.0012452
binary classification 0.001243927
positive test 0.001243759
unlabeled test 0.001241599
final classifier 0.001239904
bayesian classifier 0.001239269
small set 0.001238672
spam classification 0.001237471
different category 0.0012320360000000002
different approach 0.001221187
spy set 0.001218775
tive class 0.001209962
particular class 0.001201417
