feature weight 0.002195734
feature weights 0.002147147
feature vector 0.002138316
feature representation 0.00204164
appropriate feature 0.0019733529999999997
critical feature 0.0019671239999999998
action label 0.001882035
feature 0.00172687
same action 0.0016020969999999998
first action 0.001479172
binary features 0.001463943
learning task 0.001444702
tion label 0.001437098
viterbi algorithm 0.001414829
label sequences 0.001391765
training set 0.0013675839999999998
baseline algorithm 0.0013350900000000001
graphical model 0.001303
such models 0.0012740450000000001
action labels 0.001250716
second action 0.001222336
next action 0.0011373199999999998
features 0.00112531
same landmark 0.001121406
action chunks 0.001091718
label 0.00108989
input sequence 0.001061122
action detection 0.001047536
action detec 0.001032306
action steps 0.001032306
algorithm 0.0010203
first door 9.97255E-4
complex models 9.77353E-4
first column 9.728650000000001E-4
learning 9.69647E-4
first person 9.672190000000001E-4
different route 9.63305E-4
model 9.57786E-4
first left 9.54181E-4
first hallway 9.36987E-4
possible labeling 9.02247E-4
average number 9.02012E-4
natural language 8.90881E-4
weight vector 8.8031E-4
clean data 8.67977E-4
extra information 8.62977E-4
information extraction 8.4478E-4
problem formulation 8.42052E-4
processing tasks 8.38991E-4
token node 8.37477E-4
above problem 8.354389999999999E-4
different episodes 8.33321E-4
performance statistics 8.23229E-4
strong domain 8.18279E-4
output labels 8.17751E-4
total number 8.153839999999999E-4
shared task 8.121199999999999E-4
conditional random 8.10211E-4
average length 8.09266E-4
ordinal number 8.026859999999999E-4
possible actions 8.01588E-4
dinal number 7.96107E-4
action 7.92145E-4
node part 7.90813E-4
semantic analysis 7.774629999999999E-4
noun phrase 7.737359999999999E-4
previous efforts 7.683080000000001E-4
function 7.64488E-4
task section 7.63977E-4
surface structure 7.61869E-4
back end 7.57244E-4
language interface 7.56077E-4
human sub 7.48781E-4
whole task 7.47759E-4
text processing 7.47398E-4
annotated labels 7.44291E-4
second door 7.40419E-4
sequence 7.34419E-4
last example 7.33058E-4
end computation 7.29022E-4
few sentences 7.25716E-4
term part 7.17195E-4
end posi 7.11634E-4
front end 7.095140000000001E-4
transition node 7.087969999999999E-4
baseline algo 7.072420000000001E-4
right destination 7.06908E-4
turn right 7.064630000000001E-4
segment token 7.05529E-4
start position 7.03877E-4
sentence boundary 7.002099999999999E-4
deep semantic 6.966069999999999E-4
training 6.88795E-4
single part 6.88317E-4
first 6.87027E-4
language instructions 6.86462E-4
exact match 6.81764E-4
current position 6.81658E-4
second hallway 6.80151E-4
ural language 6.737080000000001E-4
