| C04-1147 | estimation of a4 is obtained using the | Maximum Likelihood Estimator | for the geometric distri - bution |
| I05-2034 | Equation 1 are estimated by the | maximum likelihood estimator | ( MLE ) using relative frequencies |
| H89-2013 | are considerably better than the | Maximum Likelihood Estimator | ( MLE ) : r * = r . The main |
| D13-1143 | − 1 ) ( wi ) -RSB- i = 1 The | maximum likelihood estimator | for the probability is once again |
| H90-1057 | phrases in category c . This is the | maximum likelihood estimator | of the probability that a randomly |
| D10-1011 | priors is however applied to the | maximum likelihood estimator | to compensate for data sparseness |
| C96-1003 | against . that of one based on the | Maximum Likelihood Estimator | ( MLE , for short ) . We found |
| J01-1001 | into probabilities , using the | maximum likelihood estimator | ( MLE ) , the Good-Turing method |
| E14-1066 | co-occuring with unit u ' . We use | maximum likelihood estimators | . To avoid issues with degenerate |
| D13-1143 | ) closest ancestors of w . The | maximum likelihood estimator | for this probability is : ffwi |
| H05-1110 | directly from LB and LA , by using | maximum likelihood estimators | : PA ( x ) = summationtext i |
| D13-1143 | parsed into dependency trees , the | maximum likelihood estimator | for the probability P -LSB- wi |
| E97-1006 | to efficiently approximate the | maximum likelihood estimator | of P ( kj lci ) . We employ here |
| H89-2013 | simple method , known as the " | maximum likelihood estimator | " ( MLE ) , is unsuitable because |
| D12-1001 | second we can estimate using the | maximum likelihood estimator | over our source language training |
| J11-4008 | for frequent events because the | maximum likelihood estimator | is appropriate in these cases |
| C96-1003 | the MDL Principle against the | Maximum Likelihood Estimator | in word clustering , and found |
| D08-1036 | computational linguistics . A | Maximum Likelihood estimator | sets the parameters to the value |
| J02-1005 | and consistent . For example , | maximum likelihood estimators | are unbiased and consistent across |
| E97-1006 | sequence w1 • • wN , the | maximum likelihood estimator | of 9 is defined as the value |
