A Distributional Perspective on Reinforcement Learning
傳統的強化學習算法例如Q-learning算法學習的是state-action值函數，而這篇文章的核心是學習state-action的概率分佈。 具體各簡單的例子：例如我們在上班是需要經過6站地鐵，每站地鐵平均需要5分鐘，則上上班需要30分鐘。如果每個星期（5天），地鐵都會出毛病，則就需要耽誤耽擱一個小時。正常情況下上班的時間期望，也就是均值是30分鐘，在火車出現故障的情況下，則每天的上班時間期望
相關文章
- 1. Machine Learning A Probabilistic Perspective 1.Introduction
- 2. A thorough understanding of on-policy and off-policy in Reinforcement learning
- 3. Differentially Private Federated Learning: A Client Level Perspective
- 4. 【資源】Machine Learning A Bayesian_and Optimization Perspective(MLBOP)
- 5. Building a ROS simulation platform for Deep Reinforcement Learning research ...
- 6. A Survey on Transfer Learning
- 7. Introduction to Reinforcement Learning
- 8. A Unified Game-Theoretic Approach to Multi-agent Reinforcement Learning
- 9. Playing Atari with Deep Reinforcement Learning
- 10. Reinforcement Learning Exercise 4.1
- 更多相關文章...
- • Eclipse 視圖 - Eclipse 教程
- • Web 品質- 可讀性 - 網站品質教程
- • Java Agent入門實戰（一）-Instrumentation介紹與使用
- • Java Agent入門實戰（三）-JVM Attach原理與使用
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. Machine Learning A Probabilistic Perspective 1.Introduction
- 2. A thorough understanding of on-policy and off-policy in Reinforcement learning
- 3. Differentially Private Federated Learning: A Client Level Perspective
- 4. 【資源】Machine Learning A Bayesian_and Optimization Perspective(MLBOP)
- 5. Building a ROS simulation platform for Deep Reinforcement Learning research ...
- 6. A Survey on Transfer Learning
- 7. Introduction to Reinforcement Learning
- 8. A Unified Game-Theoretic Approach to Multi-agent Reinforcement Learning
- 9. Playing Atari with Deep Reinforcement Learning
- 10. Reinforcement Learning Exercise 4.1