# Transfer Learning in Deep Reinforcement Learning: A Survey

**Authors:** Zhuangdi Zhu, Kaixiang Lin, Anil K. Jain, Jiayu Zhou  
**arXiv:** [2009.07888](https://arxiv.org/abs/2009.07888)  
**Date:** 2020-09-16  
**Summary page:** https://www.summarizepaper.com/en/paper/2009.07888v7/

## Key points

- The paper provides a comprehensive overview of transfer learning approaches in deep reinforcement learning
- The authors investigate recent progress and challenges in transfer learning techniques within deep reinforcement learning
- A framework for categorizing state-of-the-art approaches is presented, along with analysis of their goals, methodologies, compatible reinforcement learning backbones, and practical applications
- Promising prospects for reinforcement learning in domains such as robotics and game-playing through knowledge transfer from external sources are highlighted
- Potential challenges that require further research to advance transfer learning techniques are identified
- The survey serves as a valuable resource for researchers and practitioners interested in understanding and applying transfer learning approaches across various domains.

## Summary

The paper "Transfer Learning in Deep Reinforcement Learning: A Survey" by Zhuangdi Zhu, Kaixiang Lin, Anil K. Jain, and Jiayu Zhou provides a comprehensive overview of transfer learning approaches in the context of deep reinforcement learning. The authors systematically investigate recent progress and challenges in transfer learning techniques within deep reinforcement learning. They present a framework for categorizing state-of-the-art approaches and analyze their goals, methodologies, compatible reinforcement learning backbones, and practical applications. The paper highlights promising prospects for reinforcement learning in domains such as robotics and game-playing through knowledge transfer from external sources. It also identifies potential challenges that require further research to advance transfer learning techniques. Overall, this survey serves as a valuable resource for researchers and practitioners interested in understanding and applying transfer learning approaches across various domains.