Goal-oriented Trajectories for Efficient Exploration

2018-07-05 · Fabio Pardo, Vitaly Levdik, Petar Kormushev

Exploration is a difficult challenge in reinforcement learning and even recent state-of-the art curiosity-based methods rely on the simple epsilon-greedy strategy to generate novelty. We argue that pure random walks do not succeed to properly expand the exploration area in most environments and propose to replace single random action choices by random goals selection followed by several steps in their direction. This approach is compatible with any curiosity-based exploration and off-policy reinforcement learning agents and generates longer and safer trajectories than individual random actions. To illustrate this, we present a task-independent agent that learns to reach coordinates in screen frames and demonstrate its ability to explore with the game Super Mario Bros. improving significantly the score of a baseline DQN agent.

📄 PDF Abstract BibTeX arXiv:1807.02078

Code (0)

등록된 구현이 없습니다.

Tasks

Efficient Explorationreinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

Learning Goal-Oriented Vision-and-Language Navigation with Self-Improving Demonstrations at Scale

2025-09-29 · Songze Li, Zun Wang, Gengze Zhou, Jialu Li 외 arxiv

Goal-oriented vision-language navigation requires robust exploration capabilities for agents to navigate to specified goals in unknown environments without step-by-step instructions. Existing methods tend to exclusively …

Vision-Language Navigation

Graph-Enhanced Exploration for Goal-oriented Reinforcement Learning

2021-09-29 · ICLR 2022 4 · Jiarui Jin, Sijin Zhou, Weinan Zhang, Tong He 외

Goal-oriented Reinforcement Learning (GoRL) is a promising approach for scaling up RL techniques on sparse reward environments requiring long horizon planning. Recent works attempt to build suitable abstraction graph of …

continuous-control Continuous Control graph construction reinforcement-learning+2

Regioned Episodic Reinforcement Learning

2021-01-01 · Jiarui Jin, Cong Chen, Ming Zhou, Weinan Zhang 외

Goal-oriented reinforcement learning algorithms are often good at exploration, not exploitation, while episodic algorithms excel at exploitation, not exploration. As a result, neither of these approaches alone can lead t…

reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

Learning from Trajectories via Subgoal Discovery

2019-11-03 · NeurIPS 2019 12 · Sujoy Paul, Jeroen van Baar, Amit K. Roy-Chowdhury

Learning to solve complex goal-oriented tasks with sparse terminal-only rewards often requires an enormous number of samples. In such cases, using a set of expert trajectories could help to learn faster. However, Imitati…

Imitation Learning Reinforcement Learning Reinforcement Learning (RL)

Object Goal Navigation using Goal-Oriented Semantic Exploration

2020-07-01 · NeurIPS 2020 12 · Devendra Singh Chaplot, Dhiraj Gandhi, Abhinav Gupta, Ruslan Salakhutdinov

This work studies the problem of object goal navigation which involves navigating to an instance of the given object category in unseen environments. End-to-end learning-based navigation methods struggle at this task as …

Object Robot Navigation