Introduction to Machine Learning

This page collects the slides and the exercise notebooks of the course. The course consists of nine lectures and four exercise sessions of two hours each, and the exercise sessions follow lectures 2, 4, 6 and 8. Each session is added here in the week in which it is taught.

Lectures

The slides open in the browser. Pressing Esc shows an overview of the deck. To save a deck as a PDF, add ?print-pdf to its address and print the page from Chrome.

1. Introduction and linear regression
2. Polynomial regression and regularisation
3. Classification and logistic regression
4. Decision trees and ensembles
5. Neural networks 1
6. Neural networks 2
7. Convolutional neural networks
8. Recurrent neural networks
9. Unsupervised learning

Repository

All released material is also kept in the public repository gitlab.kuleuven.be/u0129795/ml2627. It holds the slides, the notebooks, the course package mlcourse and the datasets of the exercises. A copy of the repository works offline, since each deck at slides/lecture-NN/index.html opens directly in a browser. The repository is downloaded as a zip archive from its page, or cloned with git:

git clone https://gitlab.kuleuven.be/u0129795/ml2627.git

Students who cannot use Colab can run the notebooks on their own computer. The README of the repository describes the conda environment for this.

Reading

The course follows three books, all free to read online. The last slide of each lecture lists the chapters that match it.

• An Introduction to Statistical Learning with Applications in Python, by James, Witten, Hastie, Tibshirani and Taylor, for lectures 1 to 4 and 9.
• The Hundred-Page Machine Learning Book, by Burkov, a short pass over the whole course.
• Understanding Deep Learning, by Prince, for the neural network lectures 5 to 8.