Guided Cost Learning: Deep Inverse Optimal Control via Policy Optimization
This work addresses the problem of defining effective cost functions in reinforcement learning for robotics, offering incremental improvements in sample efficiency and task complexity.
The paper tackled the challenge of learning cost functions from demonstrations for high-dimensional robotic systems via inverse optimal control, presenting a method that learns nonlinear costs without feature engineering and improves sample efficiency, achieving substantial gains over prior methods in simulated and real-world tasks.
Reinforcement learning can acquire complex behaviors from high-level specifications. However, defining a cost function that can be optimized effectively and encodes the correct task is challenging in practice. We explore how inverse optimal control (IOC) can be used to learn behaviors from demonstrations, with applications to torque control of high-dimensional robotic systems. Our method addresses two key challenges in inverse optimal control: first, the need for informative features and effective regularization to impose structure on the cost, and second, the difficulty of learning the cost function under unknown dynamics for high-dimensional continuous systems. To address the former challenge, we present an algorithm capable of learning arbitrary nonlinear cost functions, such as neural networks, without meticulous feature engineering. To address the latter challenge, we formulate an efficient sample-based approximation for MaxEnt IOC. We evaluate our method on a series of simulated tasks and real-world robotic manipulation problems, demonstrating substantial improvement over prior methods both in terms of task complexity and sample efficiency.