Transfer in Reinforcement Learning: A Framework and a Survey

By Alessandro Lazaric. 2012

316 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Control and Systems Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Control and Systems Engineering

Citations in the corpus

Reinforcement Learning in Healthcare: A Survey (2019, neutral): “increasing the final convergence level [45].”
Successor Features for Transfer in Reinforcement Learning (2016, background): “This process of leveraging knowledge acquired in one task to improve performance on other tasks is called transfer [25, 11].”
Source Task Creation for Curriculum Learning (2016, background): “Transfer learning [10] [27] is a recent area of research that has been shown to speed up learning on a complex task by transferring knowledge from one or more easier source tasks.”
Safe Policy Search for Lifelong Reinforcement Learning with Sublinear Regret (2015, background): “When data is in limited supply, transfer learning can significantly improve model performance on new tasks by reusing previous learned knowledge during training (Taylor & Stone, 2009; Gheshlaghi Azar et al., 2013; Lazaric, 2011; Ferrante et al., 2008; Bou Ammar et al., 2012).”
Fast Adaptation to New Environments via Policy-Dynamics Value Functions (2020, background): “By simultaneously learning about different tasks, the agent can exploit their common structure, which can lead to faster learning and better generalization to unseen tasks from the same family (Taylor & Stone, 2009; Lazaric, 2012; Ammar et al., 2012; 2014; Parisotto et al., 2015; Borsa et al., 2016;”
DREAM Architecture: a Developmental Approach to Open-Ended Learning in Robotics (2020, unknown): “avoid starting from scratch for each new task and to transfer the previously acquired knowledge to a new context [6, 7, 8, 9, 10].”
Generalized Hidden Parameter MDPs Transferable Model-based RL in a Handful of Trials (2020, background): “Transfer learning and learning transferable agents has a long history in reinforcement learning; see (Taylor and Stone 2009; Lazaric 2012) for a survey.”
Learning Curriculum Policies for Reinforcement Learning (2018, background): “Thisideaisthebasisfortransferlearning[12,25].”
Transfer Learning in Deep Reinforcement Learning: A Survey (2020, methodology): “As pioneer surveys for TL in RL, neither [11] nor [12] covered recent research over the last decade.”
Sequential Transfer in Multi-armed Bandit with Finite Set of Models (2013, background): “Recently, multi-task and transfer learning received much attention in the supervised and reinforcement learning (RL) setting with both empirical and theoretical encouraging results (see recent surveys by Pan and Yang, 2010; Lazaric, 2011).”
Sharing Knowledge in Multi-Task Deep Reinforcement Learning (2020, background): “The same benefits of extraction and exploitation of common features among the tasks achieved in MTL, can be obtained in Multi-Task Reinforcement Learning (MTRL) when training a single agent on multiple Reinforcement Learning (RL) problems with common structures (Taylor & Stone, 2009; Lazaric, 2012).”
Multi-Task Deep Reinforcement Learning for Continuous Action Control (2017, background): “[Lazaric, 2012] is a good collection of these methods.”
Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement (2018, background): “In this paper we are interested in the problem of transfer in RL (Taylor & Stone, 2009; Lazaric, 2012).”
Online Transfer Learning in Reinforcement Learning Domains (2015, methodology): “Our online transfer learning framework can be treated as an online extension of the transfer learning framework Lazaric (2012).”
Deep Unsupervised Domain Adaptation: A Review of Recent Advances and Perspectives (2022, background): “, 2020b; Wilson and Cook, 2020; Ramponi and Plank, 2020), and the broader problem of transfer learning (Pan and Yang, 2009; Cook et al., 2013; Lazaric, 2012; Shao et al., 2014; Tan et al., 2018; Zhang et al., 2019b).”
Sparse multi-task reinforcement learning (2014, unknown): “single-task learning (we refer to [26] and [15] for a comprehensive review of the more general setting of transfer RL).”