language model 0.0019728700000000003
tree data 0.001819332
learning algorithm 0.0016368720000000002
model order 0.001596757
algorithm time 0.0015537840000000001
language models 0.001551623
hierarchical model 0.001530773
model complexity 0.0015245950000000001
training algorithm 0.00147761
model definition 0.0014697030000000002
unweighted model 0.0014488020000000001
model orders 0.0014368170000000002
model scales 0.0014368170000000002
optimization algorithm 0.0014224040000000002
projection algorithm 0.001398181
other models 0.001375788
tree representation 0.001346269
suffix tree 0.001342325
feature size 0.0013391850000000001
tree structure 0.001335101
weighted models 0.001326287
gradient algorithm 0.001316527
finding algorithm 0.001307837
tree norms 0.001307466
weighted features 0.001303996
training data 0.001290519
stochastic algorithm 0.0012819200000000002
feature vector 0.001272649
powerful algorithm 0.001264317
structured tree 0.0012624119999999999
linear language 0.001258795
collapsed tree 0.001248412
tree representa 0.0012386539999999998
feature values 0.001235105
uncollapsed tree 0.001235005
fix tree 0.0012306939999999998
tree depth 0.001230639
tree encoding 0.001228554
tree depths 0.001228554
other words 0.001216605
efficient data 0.001214965
model 0.00119551
data structure 0.001179607
training corpus 0.001154992
text corpus 0.001140641
language processing 0.0011335100000000001
heap data 0.00113004
language sentences 0.001128462
finding method 0.001122977
natural language 0.001119921
such penalties 0.001094924
data subsets 0.001088855
language mod 0.001086516
other case 0.0010842970000000001
weight parameter 0.001077839
single node 0.001074073
appropriate feature 0.001065142
same values 0.001061853
same value 0.00105953
word contexts 0.001054941
groups features 0.001052676
feature weighting 0.001046047
guage models 0.001045271
language modelling 0.001041125
proximal step 0.00104006
regression models 0.001038798
regularization parameter 0.001037562
penalty function 0.001036343
algorithm 0.00101901
ural language 0.00101839
linear time 0.0010162090000000001
time linear 0.0010162090000000001
constant time 0.001013996
same order 0.0010097309999999998
efficient learning 0.001000908
objective function 9.96318E-4
perplexity values 9.93315E-4
projection step 9.61093E-4
different methods 9.547119999999999E-4
single value 9.48837E-4
proximal methods 9.45566E-4
weighted penalties 9.450540000000001E-4
such degradation 9.4387E-4
same context 9.39584E-4
penalties standard 9.3841E-4
same complexity 9.37569E-4
constant value 9.302679999999999E-4
step com 9.22628E-4
different values 9.20653E-4
smooth function 9.20539E-4
efficient time 9.1782E-4
ing methods 9.16577E-4
average perplexity 9.02569E-4
optimization algorithms 8.994809999999999E-4
average time 8.973970000000001E-4
glish words 8.9462E-4
learning algo 8.8956E-4
gradient step 8.794390000000001E-4
possible contexts 8.78148E-4
single projection 8.76962E-4
