| J08-4002 | using supervised learning with | linear function approximation | . This " pure SL " policy simply |
| J08-4002 | fold was then used to train a | linear function approximation | user model , which was used to |
| P08-1073 | State space discretisation We use | linear function approximation | in order to learn with large |
| J11-1006 | State Space Discretization We use | linear function approximation | in order to learn with large |
| J08-4002 | which we have data . We also use | linear function approximation | to address the need to generalize |
| J08-4002 | state -- action space ( also using | linear function approximation | ) . Our hybrid policy uses SL |
| J08-4002 | for 2,000 dialogues against a | linear function approximation | user model trained on the opposite |
| J08-4002 | . This is the first time that | linear function approximation | has been used for learning dialogue |
| J11-1006 | database hits . Techniques such as | linear function approximation | are useful for handling the resulting |
| J08-4002 | future reward . We claim that | linear function approximation | is an effective way to generalize |
| P06-1024 | generalisation method such as | linear function approximation | ( Henderson et al. , 2005 ) . |
| J08-4002 | covered by the data set , and the | linear function approximation | is used to handle the very large |
| E09-1078 | known SARSA algorithm , using | linear function approximation | ( Sutton and Barto , 1998 ) . |
| J08-4002 | simplest and most efficient is | linear function approximation | . Table 2 The weights used to |
| J08-4002 | example of the limitations of | linear function approximation | , and our dependence on the previous |
| N06-1035 | addressed state features by using | linear function approximation | to deal with large state spaces |
| J08-4002 | discussed in Section 2.3 , using | linear function approximation | and a normalized exponential |
| J14-4006 | reinforcement learning algorithm ( with | linear function approximation | ) ( Shapiro and Langley 2002 |
| J11-1006 | with large state-action spaces . | Linear function approximation | learns linear estimates for expected |
| J08-4002 | challenging task indicates that | linear function approximation | is a viable approach to the very |
