EnsembleDAgger: A Bayesian Approach to Safe Imitation Learning
This addresses safety concerns in imitation learning for robotics, though it is an incremental improvement over existing DAgger variants.
The paper tackles the problem of safety in imitation learning by introducing EnsembleDAgger, a probabilistic extension to DAgger that uses an ensemble of neural networks to approximate a Gaussian Process and quantify confidence as a proxy for safety. It demonstrates improved safety and learning performance compared to other methods on tasks like an inverted pendulum and MuJoCo HalfCheetah.
While imitation learning is often used in robotics, the approach frequently suffers from data mismatch and compounding errors. DAgger is an iterative algorithm that addresses these issues by aggregating training data from both the expert and novice policies, but does not consider the impact of safety. We present a probabilistic extension to DAgger, which attempts to quantify the confidence of the novice policy as a proxy for safety. Our method, EnsembleDAgger, approximates a Gaussian Process using an ensemble of neural networks. Using the variance as a measure of confidence, we compute a decision rule that captures how much we doubt the novice, thus determining when it is safe to allow the novice to act. With this approach, we aim to maximize the novice's share of actions, while constraining the probability of failure. We demonstrate improved safety and learning performance compared to other DAgger variants and classic imitation learning on an inverted pendulum and in the MuJoCo HalfCheetah environment.