本課程學習各種強化學習技術和方法，並學習如何將AI人工智慧集成到各種遊戲與實際應用專案。本課程從基礎開始，學習馬爾可夫決策過程、Actor Critic方法、策略梯度算法、DQN等基礎方法。接著學習進階增強學習方法，像是A3C、PPO、分散式RL、以及exploration等。本課程案例使用Unity ML-Agent、openai gym 等平台做為學習環境，案例研究包括Atari遊戲、2048遊戲、自駕賽車、AlphaGo、投資分析、電網排程等主題。This course learns various reinforcement learning techniques and methods, and learns how to integrate AI artificial intelligence into various games and practical application projects. This course starts from the basics and learns basic methods such as Markov decision process, Actor Critic method, policy gradient algorithm, and DQN. Then learn advanced reinforcement learning methods, such as A3C, PPO, distributed RL, and exploration, etc. This course case uses Unity ML-Agent, openai gym and other platforms as the learning environment. Case studies include Atari games, 2048 games, self-driving racing, AlphaGo, investment analysis, power grid scheduling and other topics.
書名 Deep Reinforcement Learning Hands-On
出版社 Packt Publishing
作者 Maxim Lapan
出版年 2018
Book title Deep Reinforcement Learning Hands-On
Publisher Packt Publishing
Author Maxim Lapan
Year of publication 2018
| 評分項目 Grading Method | 配分比例 Grading percentage | 說明 Description |
|---|---|---|
| 期中專題期中專題 Midterm topic |
20 | |
| 期末專題期末專題 Final topic |
30 | |
| 作業作業 Homework |
30 | |
| 課堂出席課堂出席 class attendance |
20 |
