kernel function 0.003927242
linear kernel 0.003730301
kernel methods 0.003602143
kernel mapping 0.003599622
polynomial kernel 0.003587597
kernel svm 0.003586878
kernel types 0.003512893
kernel computing 0.003498073
efficient kernel 0.003428564
kernel degree 0.003418217
kernel svms 0.003403394
kernel com 0.003389454
mial kernel 0.003373281
different feature 0.003081472
kernel 0.00308064
feature vector 0.003075841
feature vectors 0.002966111
original feature 0.002964524
frequent feature 0.0029416670000000002
new feature 0.0028589490000000004
feature combinations 0.0027802020000000003
associate feature 0.002747533
feature numbers 0.002742449
feature combination 0.002727886
feature com 0.0027215340000000003
feature combina 0.0027097140000000002
feature conjunctions 0.00270616
feature expan 0.002702318
feature fmax 0.0027017910000000003
ordered feature 0.0027017910000000003
feature combi 0.0027017910000000003
unordered feature 0.0027017910000000003
feature 0.00241272
frequent features 0.001998667
new features 0.0019159490000000001
single features 0.001852218
tree representation 0.00179648
classification time 0.0015461390000000002
tree growing 0.001531911
training time 0.001529248
time training 0.001529248
data mining 0.001518744
features 0.00146972
many classification 0.0014575220000000002
classification method 0.001454341
training set 0.001448712
time performance 0.001438516
time linear 0.001437349
linear time 0.001437349
mining time 0.001404542
mapping function 0.001365584
time testing 0.0013546460000000001
testing time 0.0013546460000000001
natural language 0.001335366
parsing task 0.001320867
mining method 0.001312744
language processing 0.001297732
learning algorithm 0.001296867
results table 0.001291549
efficient data 0.0012498140000000001
svm training 0.001247798
ture space 0.001245922
many nlp 0.001241039
support vector 0.001223017
classification tasks 0.001211192
training examples 0.001207502
polynomial kernels 0.001205585
new classification 0.00120468
classification problem 0.001191271
training example 0.001187177
training error 0.00117965
weight vector 0.001177635
other words 0.0011572219999999999
mining approach 0.001153647
ing time 0.001151905
loss function 0.00114962
same pattern 0.001143659
new method 0.001142119
time costs 0.001127439
svm algorithm 0.001126601
experimental results 0.00112643
time purpose 0.001123007
training errors 0.001121604
machine learning 0.001118323
support vectors 0.0011132870000000001
linear form 0.001106873
training instance 0.001104127
approximate method 0.001101865
time com 0.001096502
linear ker 0.001095803
real time 0.001092923
time complexity 0.001086032
other hand 0.001084691
natural linear 0.001084434
pattern mining 0.001077847
svm testing 0.001073196
root nodes 0.001066213
mental results 0.001059632
testing speed 0.001057859
vector machines 0.001056663
