methodology-7

Transfer Reinforcement Learning

3260 papers • 126 benchmarks • 313 datasets

This task has no description! Would you like to contribute one?

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in transfer-reinforcement-learning-15

Trend

Dataset

Best Model

Actions

No benchmarks available.

Libraries

i

Use these libraries to find transfer-reinforcement-learning-15 models and implementations

Datasets

bipedal-skills

Subtasks

No subtasks available.

Most implemented papers

Actor-Mimic: Deep Multitask and Transfer Reinforcement Learning

R. Salakhutdinov, Emilio Parisotto, Jimmy Ba•Wed Nov 18 2015

This work defines a novel method of multitask and transfer learning that enables an autonomous agent to learn how to behave in multiple tasks simultaneously, and then generalize its knowledge to new domains, and uses Atari games as a testing environment to demonstrate these methods.

625

Content

0

Paper Graph

MULTIPOLAR: Multi-Source Policy Aggregation for Transfer Reinforcement Learning between Diverse Environmental Dynamics

Ryo Yonetani, M. Barekatain, Masashi Hamaya•Fri Sep 27 2019

This work explores a new challenge in transfer RL, where only a set of source policies collected under diverse unknown dynamics is available for learning a target task efficiently, and proposes MULTI-source POLicy AggRegation (MULTIPOLAR), which learns to aggregate the actions provided by the source policies adaptively to maximize the target task performance.

27 0

Paper Graph

Shapechanger: Environments for Transfer Learning

Sébastien M. R. Arnold, Tsam Kiu Pun, Théo-Tim J. Denisart, F. Cuevas•Thu Sep 14 2017

Shapechanger, a library for transfer reinforcement learning specifically designed for robotic tasks, considers three types of knowledge transfer---from simulation to simulation, from simulation to real, and from real to real---and a wide range of tasks with continuous states and actions.

0 0

Paper Graph

Universal Planning Networks

Chelsea Finn, P. Abbeel, S. Levine, A. Srinivas, A. Jabri•Sun Apr 01 2018

This work finds that the representations learned are not only effective for goal-directed visual imitation via gradient-based trajectory optimization, but can also provide a metric for specifying goals using images.

146 0

Paper Graph

Deep Transfer Reinforcement Learning for Text Summarization

Chandan K. Reddy, Naren Ramakrishnan, Yaser Keneshloo•Sun Oct 14 2018

This work proposes a reinforcement learning framework based on a self-critic policy gradient approach which achieves good generalization and state-of-the-art results on a variety of datasets and provides a generic solution that works well on unseen data.

38 0

Paper Graph

gym-gazebo2, a toolkit for reinforcement learning using ROS 2 and Gazebo

N. G. Lopez, Y. Nuin, Elias Barba Moral, Lander Usategui San Juan, A. Rueda, V. Vilches, R. Kojcev•Wed Mar 13 2019

An upgraded, real world application oriented version of gym-gazebo, the Robot Operating System (ROS) and Gazebo based Reinforcement Learning (RL) toolkit, which complies with OpenAI Gym is presented.

39 0

Paper Graph

Hardware Conditioned Policies for Multi-Robot Transfer Learning

A. Gupta, Tao Chen, Adithyavairavan Murali•Fri Nov 23 2018

This work uses the kinematic structure directly as the hardware encoding and shows great zero-shot transfer to completely novel robots not seen during training and demonstrates that fine-tuning the policy network is significantly more sample-efficient than training a model from scratch.

116 0

Paper Graph

Action Priors for Large Action Spaces in Robotics

Ondrej Biza, Dian Wang, Robert W. Platt, Jan-Willem van de Meent, Lawson L. S. Wong•Sun Jan 10 2021

The results indicate that this approach can be used to solve robotic manipulation problems that would otherwise be infeasible without expert demonstrations.

15 0

Paper Graph

Domain Adaptation In Reinforcement Learning Via Latent Unified State Representation

J. Krichmar, Xinyun Zou, Jinwei Xing, Takashi Nagata, Kexin Chen, Emre Neftci•Tue Feb 09 2021

This work proposes a two-stage RL agent that first learns a latent unified state representation (LUSR) which is consistent across multiple domains in the first stage, and then does RL training in one source domain based on LUSR in the second stage to achieve state-of-the-art domain adaptation performance.

59 0

Paper Graph

Adding a benchmark result helps the community track progress.