Learning from Data (Fall 2020)

Welcome to the class website of Learning from Data!

Announcement

2020-12-3: The guest lecture, given by Pedro Baiz, will be given from 10:00am - 11:00 am at Info Building rom 510.

2020-10-29: The midterm exam will be held in class on Friday, November 6th. For anyone who can not attend the in-class exam, please contact a TA to fill out the online exam application before Nov 3rd.

2020-09-28: The make-up lecture for the National holiday will be held on Sunday, Sept 28th at 9:20am

2020-09-09: Since the first scheduled class conflicts with the Graduate School Admission Interview, the first class is switched to Sunday morning.

Class info

  • Time: Friday 8:50-11:25

  • Location: Info Building 510

For more information about grading, homework and exam policies, see the class syllabus.

Description

This introductory course gives an overview of many concepts, techniques, and algorithms in machine learning, beginning with topics such as logistic regression and SVM and ending up with more recent topics such as deep neural networks and reinforcement learning. The course will give the student the basic ideas and intuition behind modern machine learning methods as well as a bit more formal understanding of how, why, and when they work. The underlying theme in the course is statistical inference as it provides the foundation for most of the methods covered.

Prerequisites

Basic concepts in calculus, probability theory, and linear algebra.

Team

Instructor:

  • Yang Li <yangli@sz.tsinghua.edu.cn>

TAs

  • Weida Wang <wangwd19@mails.tsinghua.edu.cn>

  • Feng Zhao <zhaof17@mails.tsinghua.edu.cn>

Office Hour

  • Professor Yang: Friday 2:00-4:00pm

  • TA Feng Zhao: Friday 8:00-10:00pm

  • TA Weida Wang: Wednesday 6:00-8:00pm

Schedule

Date Topic Homework release
9/18 Review Session (optional)
(notes) | (slides) | (code snippets)

WA0 | WA0 solution

9/20 Sunday 9:20-12:00am Introduction (make up for 9/18) (slides)

9/25 Supervised Learning I
(slides) | (slides with notes)
PA1
9/28 Supervised Learning II:
Generalized linear model
Model selection
(slides) | (slides with notes)
WA1 | WA1 solution
10/09 Supervised Learning III:
Generative model: GDA
Generative model: naive Bayesian model
(slides) | (slides with notes) | (code snippets)
10/16 Supervised Learning IV
Support vector machines
(slides) | (slides with notes)
WA2 | WA2 solution
10/23 Supervised Learning V:
Deep neural networks (slides #1) | (slides #2)
PA2 | PA2 tutorial
10/30 Unsupervised Learning I:
K-means clustering
Principal component analysis
(slides)
11/6 Midterm Exam
11/13 Unsupervised Learning II:
Independent component analysis
Canonical component analysis
(slides)
WA3 | WA3 solution
11/20 Unsupervised Learning III:
Maximal HGR correlation
Spectral Clustering (slides)
PA3 | partial solution
11/27 Unsupervised Learning IV:
Mixture Gaussian and EM algorithm
Factor Analysis
(slides)
WA4 | WA4 solution
12/11 Reinforcement Learning:
Markov decision process
Value iteration and policy iteration
Q-Learning
(slides)
WA5 | WA5 solution
12/17 Bias and variance trade off
Model selection
Learning theory
(slides)
12/25 Advanced Topic I
Transfer Learning
(slides)