learning algorithm 0.002457577
feature vector 0.002169135
algorithm accuracy 0.002059564
first algorithm 0.00205678
new algorithm 0.00204296
contextual feature 0.002028527
ing algorithm 0.0019893109999999997
final algorithm 0.001974419
feature sets 0.001972343
times feature 0.001959757
this feature 0.001951824
type feature 0.001947701
feature type 0.001947701
spelling feature 0.0019165500000000002
unsupervised algorithm 0.001913281
particular feature 0.0019119380000000002
feature combinations 0.001897938
second algorithm 0.00188916
feature extraction 0.001882977
default feature 0.001882938
similar algorithm 0.001864442
list algorithm 0.0018585989999999998
coboost algorithm 0.0018105909999999998
adaboost algorithm 0.001796231
aboost algorithm 0.001780751
intermediate algorithm 0.001780751
feature 0.00163808
training set 0.001631663
algorithm 0.00153668
such features 0.001481055
learning task 0.0014685549999999999
learning methods 0.0014592849999999998
training data 0.0014564069999999998
different learning 0.001379314
learning problems 0.001368193
training examples 0.0013600729999999998
learning problem 0.0013301559999999999
following function 0.001328236
same method 0.0013137980000000001
learning tasks 0.001300426
weak learning 0.001299713
following features 0.001288437
possible features 0.001277602
training error 0.001273553
supervised learning 0.0012680439999999999
same label 0.001268014
first method 0.0012655090000000002
machine learning 0.0012580409999999999
same form 0.001246247
list learning 0.0012428159999999999
labeled set 0.00123099
objective function 0.001220884
data examples 0.001211034
baseline method 0.001209983
test set 0.001200996
other examples 0.0011951829999999998
general features 0.001192507
possible label 0.001189354
current set 0.001188376
other task 0.0011854909999999999
likelihood function 0.001183845
unsupervised training 0.001179324
classification case 0.0011786960000000001
contextual features 0.00117832
other methods 0.0011762209999999999
supervised set 0.0011760870000000001
learning curves 0.001165044
parameter values 0.001154158
deterministic function 0.001137436
seed set 0.001135347
such algo 0.001077972
large number 0.001077848
unsupervised case 0.0010762740000000001
soft function 0.001075959
auxiliary function 0.0010747040000000001
function zco 0.001073313
differential function 0.001071873
function sug 0.001071873
list method 0.001067328
spelling features 0.001066343
initial parameter 0.001058733
second case 0.0010521530000000001
conditional probability 0.001048157
training exam 0.001046803
accuracy accuracy 0.001045768
smoothing method 0.0010346630000000001
other classifier 0.001027135
test data 0.00102574
unlabeled data 0.001021224
ing examples 0.001009981
other side 0.001004543
method tags 0.001004347
large corpus 9.99359E-4
parameter settings 9.98208E-4
greedy method 9.96317E-4
large amount 9.88808E-4
multiclass case 9.880940000000001E-4
theoretical results 9.841939999999999E-4
large numbers 9.764610000000001E-4
case letters 9.69882E-4
