1
课程详述
COURSE SPECIFICATION
以下课程信息可能根据实际授课需要或在课程检讨之后产生变动。如对课程有任何疑问，请联
系授课教师。
The course information as follows may be subject to change, either during the session because of unforeseen
circumstances, or following review of the course at the end of the session. Queries about the course should be
directed to the course instructor.
1.
课程名称 Course Title
机器学习 Machine Learning
2.
授课院系
Originating Department
计算机科学与工程系 Department of Computer Science and Technology
3.
课程编号
Course Code
CS405
4.
课程学分 Credit Value
3
5.
课程类别
Course Type
专业选修课 Major Elective Courses
6.
授课学期
Semester
秋季 Fall
7.
授课语言
Teaching Language
中英双语 English & Chinese
8.
授课教师、所属学系、联系方
式（如属团队授课，请列明其
他授课教师）
Instructor(s), Affiliation&
Contact
（For team teaching, please list
all instructors）
郝祁，副教授，计算机科学与工程系，haoq@sustech.edu.cn
Qi Hao, Associate Professor, Department of Computer Science and Technology,
haoq@sustech.edu.cn
9.
实验员/助教、所属学系、联系
方式
Tutor/TA(s), Contact
王帅军，在读博士生，计算机科学与工程系，11849555@mail.sustc.edu.cn
Shuaijun Wang, Ph.D candidate, Department of Computer Science and Technology,
11849555@mail.sustc.edu.cn
孙垚，在读硕士生，计算机科学与工程系，11849202@mail.sustc.edu.cn
Yao Sun, Master candidate, Department of Computer Science and Technology,
11849202@mail.sustc.edu.cn
张子健，在读硕士生，计算机科学与工程系，11849337@mail.sustc.edu.cn
Zijian Zhang, Master candidate, Department of Computer Science and Technology,
11849337@mail.sustc.edu.cn
10.
选课人数限额(可不填)
2
Maximum Enrolment
（Optional）
授课方式
Delivery Method
习题/辅导/讨论
Tutorials
实验/实习
Lab/Practical
其它(请具体注明)
Other（Please specify）
总学时
Total
11.
学时数
Credit Hours
32
64
12.
先修课程、其它学习要求
Pre-requisites or Other
Academic Requirements
MA103A 线性代数 I-A Linear Algebra I-A
MA212 概率论与数理统计 Probability and Statistics
13.
后续课程、其它学习规划
Courses for which this
course is a pre-requisite
无
14.
其它要求修读本课程的学系
Cross-listing Dept.
无。接受跨系选课。
None. Applicable for other departments other than CSE.
教学大纲及教学日历 SYLLABUS
15.
教学目标 Course Objectives
本课程旨在提供有关机器学习和模式识别的基础知识，从贝叶斯方法到深度学习框架。 希望学生学习机器学习系统开发方
法，并获得基于模型和数据驱动的机器学习系统设计和集成技能。
This course aims to provide fundamental knowledge about machine learning and pattern recognition, from Bayesian
approaches to deep learning frameworks. Students are expected to learn Machine learning system development
methods, and to acquire model-based and data-driven machine learning system design and integration skills.
16.
预达学习成果 Learning Outcomes
在学习完成时，学生应该掌握线性分类器/回归，非线性分类器，特征选择方法，支持向量机，概率分布模型，神经网络，
高斯混合模型，马尔科夫决策过程，图模型，强化学习以及基于贝叶斯的深度学习模型。
Upon completion of this course, the students should master linear classifier/regression, nonlinear classifier, feature
selection method, support vector machine, probability distribution model, neural network, Gaussian mixture model,
Markov decision process, graph model, reinforcement learning and Bayesian-based deep learning model.
17.
课程内容及教学日历 （如授课语言以英文为主，则课程内容介绍可以用英文；如团队教学或模块教学，教学日历须注明
主讲人）
Course Contents (in Parts/Chapters/Sections/Weeks. Please notify name of instructor for course section(s), if
this is a team teaching or module course.)
3
第一周：导论
o 课程介绍
o 机器学习导论
[实验课] 介绍实验课软件、工具包、网络资源、参考书籍。
第二周：绪论
o 曲线拟合与正则化
o 概率与高斯分布
o 推断与决策
o 熵与信息
[实验课] 在给定数据集上，实现指定的算法，完成相关任务，可以调用库。
第三周：概率分布
o 二项分布
o 多项分布
o 高斯分布
o 指数家族
o K 最近邻算法
[实验课] 学习使用 Matlab 或者 scikit-learn 对数据进行分析和预处理。
第四周：线性回归模型
o 线性基函数模型
o 最大似然估计
o 线性回归
o 预测分布
[实验课] 用 Python 实现决策树，完成分类任务。
第五周：项目开题报告
[实验课] 开题报告。
第六周：线性分类模型
o 判别函数
o Fisher 判别
o 感知机
o 逻辑回归
o 推断与生成模型
[实验课]利用线性模型，对图像提取特征。
第七周：神经网络
4
o 前馈网络
o 反向传播
o Jacobian 矩阵与 Hessian 矩阵
o 正则化
o 贝叶斯神经网络
o 卷积神经网络与对抗生成网络
[实验课] 使用 TensorFlow 搭建一个 CNN，完成手写数字识别。
第八周：稀疏核机
o 分类支持向量机
o 回归支持向量机
o 分类相关向量机
o 回归相关向量机
[实验课] 实现 SVM，并对视频进行目标检测。
第九周：期中考试
[实验课] 期中考试
第十周：混合模型与最大期望学习
o K 均值聚类算法
o 高斯混合模型
o 最大期望算法
[实验课] 实现朴素贝叶斯算法，完成垃圾邮件分类。
第十一周：时序数据
o 隐马尔科夫模型
o 隐马尔可夫模型的最大期望算法
o 前向后向算法
o 加和乘积算法
o 维比特算法
o 卡尔曼滤波与粒子滤波
[实验课] 实现一个聚类算法，并在一张图上完成语义分割。
第十二周：图模型
o 贝叶斯网络
o 条件独立
o 马尔可夫随机场
o 图模型中的推断
[实验课] 实现 EM 算法。
5
第十三周：马尔可夫决策过程
o 动态规划
o 马尔可夫决策过程
o 部分可观察马尔可夫决策过程
o 值迭代
o 策略迭代
[实验课] 使用隐马尔科夫模型，完成对股市的预测。
第十四周：贝叶斯强化学习
o 强化学习
o Q-learning
o TD-learning
[实验课] 实现维特比算法，Baum-Welch 算法。
第十五周：贝叶斯深度学习
o 贝叶斯神经网络
o 近似推断
o Drop out
o 高斯噪声
o 深度强化学习
[实验课] 实现 RL 中值迭代与策略迭代，完成对机器人的规划。
第十六周：总结和复习
[实验课] 复习、答疑。
Week 1: Introduction
o Introduction to Course
o Introduction to Machine Learning
[Lab] Introduction to the software, tools, online resources that will be used in this module and suggested textbooks.
Week 2: Preliminary
o Curve Fitting and Regularization
o Probabilities and Gaussian Distributions
o Inference and Decision
o Entropy and Information
[Lab] Finishing classification, regression and clustering on the given dataset using Scikit-Learn libraries.
Week 3: Probability Distributions
o Binomial Distributions
6
o Multinomial Distributions
o Gaussian Distributions
o Exponential Families
o KNN
[Lab] Analyze and process data by using of Matlab or scikit-learn.
Week 4: Linear Models for Regression
o Linear Basis Function Models
o Maximum Likelihood
o Linear Regression
o Predictive Distribution
[Lab] Implement one kind of Decision Tree, to finish classifing the data.
Week 5: Proposal Presentations
[Lab] Final Project proposal.
Week 6: Linear Models for Classification
o Discriminant Functions
o Fisher Discriminant
o Perceptrons
o Logistic Regression
o Inference and Generative Models
[Lab] Using linear model to finish image feature extraction task.
Week 7: Neural Networks
o Feedforward Network
o Backpropagation
o Jacobian Matrix & Hessian Matrix
o Regularization
o Bayesian Neural Networks
o CNN and GAN
[Lab] Implement a CNN to classify minist using TensorFlow.
Week 8: Sparse Kernel Machine
o SVM for Classification
o SVM for Regression
o RVM for Classification
o RVM for Regression
[Lab] Introduction to the SVM, and use it to classify the car in the video.
Week 9: Mid-term test
[Lab] Mid-term test
7
Week 10: Mixture Models and EM Learning
o K-means Clustering
o Gaussian Mixture Model
o Expectation and Maximization
[Lab] Introduction to Bayes classifier, spam email filter.
Week 11: Sequential Data
o Hidden Markov Models
o EM for HMM
o Forward-Backward Algorithm
o Sum-Product Algorithm
o Viterbi Algorithm
o Kalman and Particle Filters
[Lab] Implement a clustering algorithm.
Week 12: Graphical Models
o Bayesian Networks
o Conditional Independence
o Markov random fields
o Inference in Graphical Models
[Lab] Implement EM algorithm with data using in lab.
Week 13: Markov Decision Process
o Dynamic programming
o Markov Decision Process
o Partially Observable MDP
o Value Iteration
o Policy Iteration
[Lab] Implement HMM algorithm.
Week 14: Bayesian Reinforcement Learning
o Reinforcement learning
o Q-learning
o TD-learning
[Lab] Implement viterbi algorithm, Baum-Welch algorithm.
Week 15: Bayesian Deep Learning
o Bayesian NN
o Approximate Inference
o Drop out
o Gaussian noise
o deep reinforcement learning
8
[Lab] Introduction to reinforcement learning, implement policy iteration and value iteration.
Week 16: Summary & Revision
[Lab] Final projects.
18.
教材及其它参考资料 Textbook and Supplementary Readings
1. Avison D, Fitzgerald G. Information systems development: methodologies, techniques and tools[M].
McGraw Hill, 2003.
2. Bishop C M. Pattern recognition and machine learning[M]. springer, 2006.
课程评估 ASSESSMENT
19.
评估形式
Type of
Assessment
评估时间
Time
占考试总成绩百分比
% of final
score
违纪处罚
Penalty
备注
Notes
出勤 Attendance
课堂表现
Class
Performance
小测验
Quiz
20%
5 次
Five times
课程项目 Projects
10%
Final project 开题报告
Final project proposal report
平时作业
Assignments
20%
8 次理论作业，8 次实验作业
8 times homeworks & lab
assignments
期中考试
Mid-Term Test
20%
开卷考试
Open-book exam
期末考试
Final Exam
20%
闭卷考试
Unseen exam
期末报告
Final
10%
Final project 结题报告
Final project report
9
Presentation
其它（可根据需要
改写以上评估方
式）
Others (The
above may be
modified as
necessary)
20.
记分方式 GRADING SYSTEM
A. 十三级等级制 Letter Grading
B. 二级记分制（通过/不通过） Pass/Fail Grading
课程审批 REVIEW AND APPROVAL
21.
本课程设置已经过以下责任人/委员会审议通过
This Course has been approved by the following person or committee of authority