59 citations · 66 across the 4 of their papers we have counts for
1 paper · 1 filter
Pierre Lison
Reinforcement learning methods are increasingly used to optimise dialogue policies from experience. Most current techniques are model-free: they directly estimate the utility of va…