Courses Detail Information

ECE4450J – Introduction to Machine Learning


Heng Qiao
Weimin Zhou

Credits: 4

Pre-requisites: ECE2810J & MATH2140J/4170J/2860J & ECE4010J


This course aims to provide an introduction to machine learning with some practical applications in R. The perspective to be taken is more statistical rather than computational. Both classic supervised and unsupervised learning techniques will be covered but more emphasis will be put on variants of regression. By taking this course, the students are supposed to acquire the basic mathematical insights behind the common machine learning methodologies. Besides, the students will learn to program in R and solve real-life problems.

Course Topics:

1. Overview of linear algebra, probability, statistics, and optimization theory
2. Linear regression, least squares, lasso
3. Logistic regression, linear discriminant analysis, Bayes’ theorem
4. Resampling methods: cross-validation, bootstrap
5. Tree-based methods: decision trees, random forests, boosting
6. Support vector machine
7. Unsupervised learning: principal component analysis, clustering methods
8. Perceptron and neural networks