CROP: Towards Distributional-Shift Robust Reinforcement Learning using Compact Reshaped Observation Processing

Abstract

The safe application of reinforcement learning (RL) requires generalization from limited training data to unseen scenarios. Yet, fulfilling tasks under changing circumstances is a key challenge in RL. Current state-of-the-art approaches for generalization apply data augmentation techniques to increase the diversity of training data. Even though this prevents overfitting to the training environment(s), it hinders policy optimization. Crafting a suitable observation, only containing crucial information, has been shown to be a challenging task itself. To improve data efficiency and generalization capabilities, we propose Compact Reshaped Observation Processing (CROP) to reduce the state information used for policy optimization. By providing only relevant information, overfitting to a specific training layout is precluded and generalization to unseen environments is improved. We formulate three CROPs that can be applied to fully observable observation- and action-spaces and provide methodical foundation. We empirically show the improvements of CROP in a distributionally shifted safety gridworld. We furthermore provide benchmark comparisons to full observability and data-augmentation in two different-sized procedurally generated mazes.

arXiv comment9 pages, 5 figures, published at IJCAI 2023

References in corpus (10)

• Generalizing from a Few Examples: A Survey on Few-Shot Learning
• OpenAI Gym
• Robust Adversarial Reinforcement Learning
• Reinforcement Learning with Augmented Data
• GridMask Data Augmentation
• Early stopping and non-parametric regression: An optimal data-dependent stopping rule
• Leveraging Procedural Generation to Benchmark Reinforcement Learning
• Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learning
• Cross-Trajectory Representation Learning for Zero-Shot Generalization in RL
• Real-time Drift Detection on Time-series Data