P14-1082 scores was tested using pairwise bootstrap sampling ( Koehn , 2004 ) . All significance
D14-1022 Significance tests are conducted using bootstrap sampling ( Koehn , 2004 ) . Our ME classifiers
W10-1757 are statistically significant by bootstrap sampling test ( 1000 samples , p <
P12-1048 . Finally , we conduct paired bootstrap sampling ( Koehn , 2004 ) to test the
P10-2003 . Finally , we conduct paired bootstrap sampling ( Koehn , 2004 ) to test the
D15-1145 Finally , we conducted paired bootstrap sampling ( Koehn , 2004 ) to test the
D15-1146 , 2002 ) . We performed paired bootstrap sampling ( Koehn , 2004 ) to test the
W07-0734 statistical significance using bootstrap sampling indicate that the differences
W11-1906 process ( Breiman , 2001 ) , first , bootstrap sampling is employed to generate multiple
W05-0908 significance levels based on the bootstrap sampling distribution , we employ the
W10-2917 all , comparing to the original bootstrap sampling initial - ization , we use a
W14-4322 statistical significance test using bootstrap sampling method given in Berg-Kirkpatrick
W99-0620 , given the training corpus . Bootstrap Sampling Procedure Repeat i = 1 K times
W05-0908 of the null hypothesis and the bootstrap sampling distribution have the same shape
Q15-1009 increase the accuracy . We did bootstrap sampling to test the statistical significance
W10-1762 statistically significant according to the bootstrap sampling test ( p < 0.05 , 1,000 samples
W10-2917 classifier is initially generated via bootstrap sampling from the original labeled training
W99-0620 method to obtain P ( M I S ) . The bootstrap sampling method provides a way for artificially
W05-0908 location . The location of the bootstrap sampling distribution is shifted so that
W10-2917 score of 0.7717 , higher than the bootstrap sampling strategy ( 0.7684 ) . This result
hide detail