Suggested work for this block

2 hours lectures, 2 hours self-study, 2 hour book reading and video watching, 2 hours computer lab session, 2 hours paper and pencil tutorial, 4 hours work on mini-project. Total 14 hours. We prepared a self-assessment quiz with true-false statements from previous exams about important concepts to help you revise.

The mini-project should now be in full swing. If you have questions regarding the mini-project, write in the discussion forum.

Lecture F6: Tree-based methods, bagging and boosting

The basics of decision trees.

Basics for bootstrap aggregation (aka. bagging)

Basics for boosting:

Lecture F6 on Friday Nov 19th @ 13.15
Recommended reading: Ch. 2.3, 7.1-7.3

Slides

Tutorial L7 (pen & paper)

Start this work after the lecture. Spend about two hours attempting the questions before the tutorial.

The exercise sheet is available here.

Recommended problems: 7.1, 7.2, 7.3, 7.4

Feel free to ask questions in the discussion section.

Tutorial 8 (computer)

Start this work after lecture. Spend about two hours attempting the questions before the session.

The computer labs are available here:

Python notebook
Google Colab

Recommended problems: 8.1, 8.2
Additional problems: 8.3

Suggested solutions are available here.

Also feel free to ask questions in the discussion section.

Tutorial 9 (computer)

Boosting can be a useful technique for those of you looking to improve your approach in the mini-project. Computer labs for boosting are available here:

Python notebook
Google Colab

Recommended problems: 9.1, 9.2, 9.3
Additional problems: 9.4, 9.5

Suggested solutions are available here.

Also feel free to ask questions in the discussion section.

A random forest model can be used to learn melodies in music and produce new music, see: