Imitation Learning with Recurrent Neural Networks

Khanh Nguyen

We present a novel view that unifies two frameworks that aim to solve sequential prediction problems: learning to search (L2S) and recurrent neural networks (RNN). We point out equ…