Maximum Entropy Deep Inverse Reinforcement Learning
This work addresses the problem of learning reward functions from demonstrations for applications in life-long learning scenarios, representing an incremental improvement over existing methods.
The paper tackles the inverse reinforcement learning problem by using neural networks to approximate complex reward functions, achieving state-of-the-art performance on existing benchmarks and exceeding on an alternative benchmark with highly varying reward structures.
This paper presents a general framework for exploiting the representational capacity of neural networks to approximate complex, nonlinear reward functions in the context of solving the inverse reinforcement learning (IRL) problem. We show in this context that the Maximum Entropy paradigm for IRL lends itself naturally to the efficient training of deep architectures. At test time, the approach leads to a computational complexity independent of the number of demonstrations, which makes it especially well-suited for applications in life-long learning scenarios. Our approach achieves performance commensurate to the state-of-the-art on existing benchmarks while exceeding on an alternative benchmark based on highly varying reward structures. Finally, we extend the basic architecture - which is equivalent to a simplified subclass of Fully Convolutional Neural Networks (FCNNs) with width one - to include larger convolutions in order to eliminate dependency on precomputed spatial features and work on raw input representations.