Evaluating the progress of Deep Reinforcement Learning in the real world: aligning domain-agnostic and domain-specific research
arXiv:2107.03015
Abstract
Deep Reinforcement Learning (DRL) is considered a potential framework to improve many real-world autonomous systems; it has attracted the attention of multiple and diverse fields. Nevertheless, the successful deployment in the real world is a test most of DRL models still need to pass. In this work we focus on this issue by reviewing and evaluating the research efforts from both domain-agnostic and domain-specific communities. On one hand, we offer a comprehensive summary of DRL challenges and summarize the different proposals to mitigate them; this helps identifying five gaps of domain-agnostic research. On the other hand, from the domain-specific perspective, we discuss different success stories and argue why other models might fail to be deployed. Finally, we take up on ways to move forward accounting for both perspectives.
References in corpus (27)
- Continuous control with deep reinforcement learning
- Overcoming catastrophic forgetting in neural networks
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
- Deep Reinforcement Learning for Multi-Agent Systems: A Review of Challenges, Solutions and Applications
- Meta-SGD: Learning to Learn Quickly for Few-Shot Learning
- Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems
- Deep learning for molecular design - a review of the state of the art
- IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
- How to Train Your Robot with Deep Reinforcement Learning; Lessons We've Learned
- Deep Reinforcement Learning: An Overview
- Counterfactual Multi-Agent Policy Gradients
- Deep Reinforcement Learning that Matters
- Safe Model-based Reinforcement Learning with Stability Guarantees
- Safe Exploration in Continuous Action Spaces
- Challenges of Real-World Reinforcement Learning
- QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement Learning
- Deep Decentralized Multi-task Multi-Agent Reinforcement Learning under Partial Observability
- MOReL : Model-Based Offline Reinforcement Learning
- Sample-Efficient Reinforcement Learning with Stochastic Ensemble Value Expansion
- Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learning
- Learning Invariant Representations for Reinforcement Learning without Reconstruction
- Self-Supervised Policy Adaptation during Deployment
- Improving Generalization in Meta Reinforcement Learning using Learned Objectives
- Keep Doing What Worked: Behavioral Modelling Priors for Offline Reinforcement Learning
- Q-Learning in enormous action spaces via amortized approximate maximization
- Reward Shaping with Recurrent Neural Networks for Speeding up On-Line Policy Learning in Spoken Dialogue Systems
- Observation Space Matters: Benchmark and Optimization Algorithm