| D09-1005 | also be used for second-order | gradient optimization | algorithms . We implement the |
| N10-1137 | results . We use a stochastic | gradient optimization | method ( Bottou , 2004 ) to optimize |
| H90-1075 | backpropagation with conjugate | gradient optimization | \ -LSB- 6 \ -RSB- . Each network |
| D10-1124 | be updated in closed form , so | gradient optimization | is again required . The derivation |
| P14-1034 | 6 are optimized using L - BFGS | gradient optimization | ( Galassi et al. , 2003 ) . We |
| S14-2012 | latter , we used the truncated | gradient optimization | ( Langford et al. , 2009 ) , |
| P10-1131 | accomplished by applying standard | gradient optimization | methods . Second , while the |
| P06-1027 | CRFs , we may apply stochastic | gradient optimization | method with adaptive gain adjustment |
| P15-1058 | regression , NER tagging and Conjugate | Gradient optimization | . For NER tagging we used a pre-trained |
| W06-3603 | parameters to zero during conjugate | gradient optimization | , which are then pruned before |
| W04-3223 | be driven to zero in conjugate | gradient optimization | of the Bl-regularized objective |
| P11-1075 | 1 , using a primal stochastic | gradient optimization | algorithm that follows Shalev |
