強化學習

Reinforcement Learning

學期

110-2

學分

0 學分

當期課號

5237

永久課號

IOG5029

開課單位

智慧科學暨綠能學院

授課教師

林勻蔚

校區

歸仁

類別

選修

上課時間表

節	週三
5 13:20–14:10	強化學習 CM218（歸仁） 3 節連堂
6 14:20–15:10
7 15:30–16:20

* 根據陽明交大上課時間表所列

概述

This course provides a clear and simple account of the key ideas and algorithms of reinforcement learning and takes the point of view of artificial intelligence and engineering. We also survey some of the frontiers of reinforcement learning in biology and applications.

先修科目

Python

備註

無備註

教學方式

教師未提供此項資料

評分方式

In-class projects (60%) Final project (40%)

課程大綱

Tabular Solution Methods Introduction

Multi-arm Bandits Finite Markov Decision Processes Dynamic Programming Monte Carlo Methods Temporal-Difference Learning Eligibility Traces Planning and Learning with Tabular Methods

講授：

16
實作：

8
Approximate Solution Methods Introduction

On-policy Approximation of Action Values Off-policy Approximation of Action Values Policy Approximation

講授：

8
實作：

4
Frontiers

Psychology Neuroscience Applications and Case Studies

講授：

8
實作：

4
The Reinforcement Learning Problem

The Reinforcement Learning Problem Introduction

講授：

6

週次計畫

週次	主題
第 1 週	The Reinforcement Learning Problem
第 2 週	Tabular Solution Methods Introduction
第 3 週	Multi-arm Bandits
第 4 週	Finite Markov Decision Processes
第 5 週	Dynamic Programming
第 6 週	Monte Carlo Methods
第 7 週	Temporal-Difference Learning
第 8 週	Eligibility Traces
第 9 週	Planning and Learning with Tabular Methods Introduction
第 10 週	Approximate Solution Methods Introduction
第 11 週	On-policy Approximation of Action Values
第 12 週	Off-policy Approximation of Action Values
第 13 週	Policy Approximation
第 14 週	Psychology
第 15 週	Neuroscience
第 16 週	Applications and Case Studies
第 17 週	Final Project Presentation
第 18 週	Final Project Demonstration

教科書

教師未提供此項資料

Office Hours

地點: Go far 209
時間: 10:00~12:00, Wed.
聯絡方式: 03-5731350