Reinforcement Learning: A Survey
Обучение с подкреплением: обзор
1996-05-01
SCID: 54.1/gffwh7nk
Discuss with AI
Markov decision theorydelayed reinforcementexploration-exploitation tradeoffhidden statereinforcement learning
Figures from the paper
Abstract (AI)
This paper surveys the field of reinforcement learning from a computer-science perspective. It is written to be accessible to researchers familiar with machine learning. Both the historical basis of the field and a broad selection of current work are summarized. Reinforcement learning is the problem faced by an agent that learns behavior through trial-and-error interactions with a dynamic environment. The work described here has a resemblance to work in psychology, but differs considerably in the details and in the use of the word ``reinforcement.'' The paper discusses central issues of reinforcement learning, including trading off exploration and exploitation, establishing the foundations of the field via Markov decision theory, learning from delayed reinforcement, constructing empirical models to accelerate learning, making use of generalization and hierarchy, and coping with hidden state. It concludes with a survey of some implemented systems and an assessment of the practical utility of current methods for reinforcement learning.
Key Findings
1
An assessment of implemented systems evaluates the practical utility of contemporary reinforcement-learning methods.
2
Key research challenges include building empirical models to accelerate learning, exploiting generalization and hierarchy, and coping with hidden state.
3
Reinforcement learning is framed as an agent learning behavior through trial-and-error interactions with a dynamic environment.
4
The field’s foundations are established through Markov decision theory, while addressing exploration–exploitation trade-offs and delayed reinforcement.
5
The survey connects reinforcement learning historically to psychology while emphasizing substantial differences in technical details and terminology.
Research Object
Reinforcement learning (agent learning behavior via trial-and-error interactions with a dynamic environment)
Research Subject
learning behavior through trial-and-error, including exploration–exploitation trade-offs, delayed reinforcement, generalization, hierarchy, and hidden-state handling
Publication Details
Publication Date
1996-05-01
Journal
Publisher
ISSN
Cited by
8968
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai2
Cited by14
Reinforcement learning in robotics: A survey2013
AI-Based Modeling: Techniques, Applications and Research Issues Towards Automation, Intelligent and Smart Systems2022
Power to the People: The Role of Humans in Interactive Machine Learning2014
Toward Causal Representation Learning2021
Active learning in materials science with emphasis on adaptive sampling using uncertainties for targeted design2019
Interactive machine learning for health informatics: when do we need the human-in-the-loop?2016
Cybersecurity data science: an overview from machine learning perspective2020
A wearable cardiac ultrasound imager2023
Machine Learning for 5G/B5G Mobile and Wireless Communications: Potential, Limitations, and Future Directions2019
Efficient Acceleration of Deep Learning Inference on Resource-Constrained Edge Devices: A Review2022
Leveraging artificial intelligence and advanced food processing techniques for enhanced food safety, quality, and security: a comprehensive review2025
Exploring Large Language Model‐Based Intelligent Agents: Definitions, Methods, and Prospects2026
RLHF Deciphered: A Critical Analysis of Reinforcement Learning from Human Feedback for LLMs2025
Agentic Artificial Intelligence for Smart Grids: A Comprehensive Review of Autonomous, Safe, and Explainable Control Frameworks2026