learning methods 0.002631395
active learning 0.002571234
learning algorithm 0.002562528
online learning 0.002357048
learning accuracy 0.002268599
learning algorithms 0.002155373
machine learning 0.002133965
learning techniques 0.002094835
learning algo 0.002081094
tive learning 0.002062288
learning setting 0.002053463
gin learning 0.002033891
confident learning 0.002022176
learning frameworks 0.002021354
learning rounds 0.002021307
learning speed 0.002019718
criminative learning 0.002018831
learning methodologies 0.002018831
training data 0.001929875
learning 0.00176954
language data 0.001719459
labeled training 0.0015534429999999998
online method 0.001508972
margin methods 0.001483909
online methods 0.001449363
single feature 0.00142278
training instances 0.001412259
test data 0.001409136
online algorithm 0.0013804960000000002
parameter confidence 0.001354655
random selection 0.001349958
many methods 0.001331812
several methods 0.001315257
margin confidence 0.001299077
online classifier 0.00129382
labeled examples 0.001293542
probabilistic methods 0.001271463
feature weights 0.001255075
standard methods 0.0012354100000000002
labeled set 0.0012149029999999998
prediction confidence 0.001205301
tial training 0.0011880229999999999
ing algorithm 0.001185943
parameter value 0.001148451
different dataset 0.001141085
ing labeling 0.001139441
large margin 0.001139171
japanese word 0.0011357440000000002
based methods 0.001132896
abilistic methods 0.001113027
natural language 0.001102898
margin value 0.001092873
selection criteria 0.0010924959999999999
similar labels 0.001092295
simple margin 0.001090781
new model 0.0010887470000000002
labeling cost 0.001085941
new parameters 0.001070343
unlabeled examples 0.001068476
correct label 0.0010635430000000001
random instance 0.001063389
active learner 0.00105697
small margin 0.0010550450000000001
active confident 0.00105433
active learners 0.0010506790000000001
selection baselines 0.001046018
word sense 0.0010458210000000002
large number 0.001045704
low confidence 0.001040934
corpus diversity 0.0010377239999999999
selection crite 0.001014814
parameter vectors 0.0010083969999999999
labeling efforts 0.001003718
margin scores 0.001001289
valuable information 9.982419999999999E-4
parameter con 9.946269999999999E-4
word segmentation 9.92033E-4
probabilistic measure 9.884009999999999E-4
nlp performance 9.85807E-4
based approach 9.8303E-4
high prediction 9.79423E-4
mean parameter 9.756119999999999E-4
small number 9.615779999999999E-4
margin constraint 9.600570000000001E-4
single instance 9.582989999999999E-4
gram features 9.56549E-4
parameter change 9.544E-4
single value 9.5194E-4
margin criteria 9.508030000000001E-4
margin values 9.499440000000001E-4
classifiers common 9.471479999999999E-4
probability prediction 9.43678E-4
feature 9.41659E-4
training 9.37514E-4
parameter confi 9.37268E-4
creases confidence 9.28992E-4
decision tasks 9.26146E-4
method 9.21464E-4
test accuracy 9.15834E-4
gaussian distribution 9.069810000000001E-4
