Artificial Intelligence
逆强化学习
Inverse Reinforcement LearningA framework for learning an agent's objective function (reward function) by observing its behavior in a complex environment, rather than being explicitly told what the rewards are. / 一种通过观察智能体在复杂环境中的行为来学习其目标函数（奖励函数）的框架，而不是被明确告知奖励是什么。
- Knowledge type
- 数学理论
- First proposed or discovered
- Proposer
- Turkey Farm member
- Creator
- 农场主
- Last updated
- 2026-11-10 20:06:55
Sub-knowledge points
Maximum Entropy Inverse Reinforcement LearningCost-based Inverse Reinforcement LearningContextual Inverse Reinforcement Learning
Higher-order knowledge points
Main parameters and performance indicators
Observation: The state or environment dataAction: The sequence of decisions made by the agentPolicy: The behavior strategy of the agent
Open problems
Sample efficiencyScalability to high-dimensional state spacesRobustness to adversarial or corrupted data
Related literature
ranked by community starsPhysics-constrained maximum entropy inverse reinforcement learning with learned-cost rollout for spacecraft pursuit evasion in elliptical orbitsOpen source
Related products and links
No item recorded yet
Creation and modification trail
1 recordsRecord title: Inverse Reinforcement Learning.
Node update feedback
review required