4,054 Hits in 3.3 sec

Automatic Curriculum Graph Generation for Reinforcement Learning Agents

Nick Walker, Rishi Shah
2017 Zenodo  
Poster presented at AAAI 2017 in San Francisco for  ...  Right: A curriculum generated from 11 input tasks in the Block Dude domain. 1 2 6 3 4 … … … … … … Programming Data Structures Discrete Maths Operating Systems • Curriculum learning  ...  target task ○ 1 2 6 3 4 Target 1 2 6 3 4 Output • Directed acyclic graph • Topological ordering gives an execution sequence Target 1 2 6 3 4 Expected reward Agent  ... 
doi:10.5281/zenodo.3244635 fatcat:4v4r4jdprrerdgoky2murbp77q

Towards Practical Multi-Object Manipulation using Relational Reinforcement Learning [article]

Richard Li, Allan Jabri, Trevor Darrell, Pulkit Agrawal
2019 arXiv   pre-print
Learning from a curriculum of increasingly complex tasks appears to be a natural solution, but unfortunately, does not work for many scenarios.  ...  Starting from scratch, our agent learns to stack six blocks into a tower.  ...  We thank Amazon Web Services (AWS) for their generous support in the form of cloud credits.  ... 
arXiv:1912.11032v1 fatcat:zumbjf6rp5g3bdjqysjiu7jhxa

Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey [article]

Sanmit Narvekar and Bei Peng and Matteo Leonetti and Jivko Sinapov and Matthew E. Taylor and Peter Stone
2020 arXiv   pre-print
Reinforcement learning (RL) is a popular paradigm for addressing sequential decision tasks in which the agent has only limited environmental feedback.  ...  In this article, we present a framework for curriculum learning (CL) in reinforcement learning, and use it to survey and classify existing CL methods in terms of their assumptions, capabilities, and goals  ...  Part of this work has taken place in the Learning Agents Research Group (LARG) at the Artificial Intelligence Laboratory, The University of Texas at Austin. LARG re-  ... 
arXiv:2003.04960v2 fatcat:iacmqeb7jjeezpo27jsnzuqb7u

Learning Object-Centered Autotelic Behaviors with Graph Neural Networks [article]

Ahmed Akakzia, Olivier Sigaud
2022 arXiv   pre-print
In this paper, we propose to investigate the impact of these representations on the learning capabilities of autotelic agents.  ...  We study different implementations of autotelic agents using four types of Graph Neural Networks policy representations and two types of goal spaces, either geometric or predicate-based.  ...  The authors would like to thank Hugo Caselles-Dupré and Mohamed Chetouani for insightful discussions.  ... 
arXiv:2204.05141v1 fatcat:pb44nkgtxzbbrk62ox2cdnoy7m

Self-Paced Multi-Agent Reinforcement Learning [article]

Wenshuai Zhao, Joni Pajarinen
2022 arXiv   pre-print
Curriculum reinforcement learning (CRL) aims to speed up learning of a task by changing gradually the difficulty of the task from easy to hard through control of factors such as initial state or environment  ...  While automating CRL is well studied in the single-agent setting, in multi-agent reinforcement learning (MARL) an open question is whether control of the number of agents with other factors in a principled  ...  Conclusion This paper presents a principled approach, self-paced multi-agent reinforcement learning (SPMARL), to generate curricula for MARL.  ... 
arXiv:2205.10016v1 fatcat:gjfziyzbh5gc3mlzrhcbddumuy

Teacher-Student Curriculum Learning [article]

Tambet Matiisen, Avital Oliver, Taco Cohen, John Schulman
2017 arXiv   pre-print
We propose Teacher-Student Curriculum Learning (TSCL), a framework for automatic curriculum learning, where the Student tries to learn a complex task and the Teacher automatically chooses subtasks from  ...  Using our automatically generated curriculum enabled to solve a Minecraft maze that could not be solved at all when training directly on solving the maze, and the learning was an order of magnitude faster  ...  Conclusion We presented a framework for automatic curriculum learning that can be used for supervised and reinforcement learning tasks.  ... 
arXiv:1707.00183v2 fatcat:uj7p2eonlnawzpm2xulkrgw5hy

Curiosity Based Reinforcement Learning on Robot Manufacturing Cell [article]

Mohammed Sharafath Abdul Hameed, Md Muzahid Khan, Andreas Schwung
2020 arXiv   pre-print
Further, the learning agents are embedded into the transportation robots to enable a generalized learning solution that can be applied to a variety of environments.  ...  Reinforcement learning has proved to be highly successful in solving tasks like robotics and scheduling.  ...  Graph network based curriculum learning: In the next experiment, the intrinsic curiosity module is plugged to train the agents on gRMC. The learning process was a lot more difficult in this setup.  ... 
arXiv:2011.08743v1 fatcat:tbwjdbpodfh2dohpecjai2lily

Deep Reinforcement Learning for Navigation in AAA Video Games [article]

Eloi Alonso, Maxim Peter, David Goumard, Joshua Romoff
2020 arXiv   pre-print
As an alternative, we propose to use Deep Reinforcement Learning (Deep RL) to learn how to navigate 3D maps using any navigation ability.  ...  The most popular approach for NPC navigation in the video game industry is to use a navigation mesh (NavMesh), which is a graph representation of the map, with nodes and edges indicating traversable areas  ...  Finally, we thank Olivier Delalleau, Batu Aytemiz and Sahand Rezaei-Shoshtari for their contribution to an early iteration of this project.  ... 
arXiv:2011.04764v2 fatcat:kr45vvg3ibbkti6xv5zosbxvs4

A Novel Automated Curriculum Strategy to Solve Hard Sokoban Planning Instances [article]

Dieqiao Feng, Carla P. Gomes, Bart Selman
2021 arXiv   pre-print
In recent years, we have witnessed tremendous progress in deep reinforcement learning (RL) for tasks such as Go, Chess, video games, and robot control.  ...  Our RL agent can solve hard instances that are far out of reach for any previous state-of-the-art Sokoban solver.  ...  Acknowledgements We thank the reviewers for valuable feedback.  ... 
arXiv:2110.00898v1 fatcat:2737p3wcvfaxvcmti5kgxh4tk4

Blocks Assemble! Learning to Assemble with Large-Scale Structured Reinforcement Learning [article]

Seyed Kamyar Seyed Ghasemipour, Daniel Freeman, Byron David, Shixiang Shane Gu, Satoshi Kataoka, Igor Mordatch
2022 arXiv   pre-print
We find that the combination of large-scale reinforcement learning and graph-based policies -- surprisingly without any additional complexity -- is an effective recipe for training agents that not only  ...  Through extensive experiments, we highlight the importance of large-scale training, structured representations, contributions of multi-task vs. single-task learning, as well as the effects of curriculums  ...  Our contributions are as follows: • We introduce an assembly domain that allows for a controlled study of generalization in reinforcement learning (RL). • We demonstrate a single agent that can simultaneously  ... 
arXiv:2203.13733v2 fatcat:67mvji62pvemtckgyzwr6ffuvy

Composable Planning with Attributes [article]

Amy Zhang, Adam Lerer, Sainbayar Sukhbaatar, Rob Fergus, Arthur Szlam
2019 arXiv   pre-print
We propose a method that learns a policy for transitioning between "nearby" sets of attributes, and maintains a graph of possible transitions.  ...  We show in 3D block stacking, grid-world games, and StarCraft that our model is able to generalize to longer, more complex tasks at test time by composing simpler learned policies.  ...  AP outperforms reinforcement learning even when a curriculum is provided.  ... 
arXiv:1803.00512v2 fatcat:wt5f7qrrljdq5pdyc7xxxr3gaa

Learning Multi-Objective Curricula for Deep Reinforcement Learning [article]

Jikun Kang, Miao Liu, Abhinav Gupta, Chris Pal, Xue Liu, Jie Fu
2022 arXiv   pre-print
Various automatic curriculum learning (ACL) methods have been proposed to improve the sample efficiency and final performance of deep reinforcement learning (DRL).  ...  For example, ACL can be used for subgoal generation, reward shaping, environment generation, or initial state generation.  ...  Automatic curriculum learning (ACL) for deep reinforcement learning (DRL) [Portelas et al., 2020a] has recently emerged as a promising tool to learn how to adapt an agent's learning tasks to its capabilities  ... 
arXiv:2110.03032v2 fatcat:u3az7o7qzvfjjb4ldfytzakqqe

CLAIM: Curriculum Learning Policy for Influence Maximization in Unknown Social Networks [article]

Dexun Li, Meghna Lowalekar, Pradeep Varakantham
2021 arXiv   pre-print
In this work, we propose CLAIM - Curriculum LeArning Policy for Influence Maximization to improve the sample efficiency of RL methods.  ...  In this paper, we focus on this problem of network discovery for influence maximization. The existing work in this direction proposes a reinforcement learning framework.  ...  A recent work by Kamarthi et al. [2019] provides a reinforcement learning based approach to automatically train an agent for network discovery.  ... 
arXiv:2107.03603v1 fatcat:vcu3j5om4jhxzmyds53lzwh2oy

Autonomous Task Sequencing for Customized Curriculum Design in Reinforcement Learning

Sanmit Narvekar, Jivko Sinapov, Peter Stone
2017 Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence  
We use our approach to automatically sequence tasks for 3 agents with varying sensing and action capabilities in an experimental domain, and show that our method produces curricula customized for each  ...  Recent work has shown that transfer learning can be extended to the idea of curriculum learning, where the agent incrementally accumulates knowledge over a sequence of tasks (i.e. a curriculum).  ...  Acknowledgements The authors would like to thank Matteo Leonetti for helpful discussions early on in this work.  ... 
doi:10.24963/ijcai.2017/353 dblp:conf/ijcai/NarvekarSS17 fatcat:lcdr7tbasrdolpvcaxpm4ki2om

Curriculum goal masking for continuous deep reinforcement learning [article]

Manfred Eppe, Sven Magg, Stefan Wermter
2019 arXiv   pre-print
In this work, we present a simple and general goal masking method that also allows us to estimate a goal's difficulty level and thus realize a curriculum learning approach for deep RL.  ...  Deep reinforcement learning has recently gained a focus on problems where policy or value functions are independent of goals.  ...  Curriculum learning for deep reinforcement learning The term curriculum learning (CL) for neural networks was coined by Bengio et al.  ... 
arXiv:1809.06146v2 fatcat:2dycfs5ofnh3jg4aza3t6azqca
« Previous Showing results 1 — 15 out of 4,054 results