Documentation of the VU Hands on Machine Learning

This is the documentation of the VU 2.0 ECTS Hands on Machine Learning course at the University of Applied Arts Vienna (short: dieAngewandte). This is the second (and still experimental) iteration of this course in summer term 2024. For the archived documentation of the first instance in 2023 go to ../2023.

The general schedule for this course (all sessions from 14:00 to 18:00):

• 11 March 2024, SR 35 : Common kick-off session with SE Machine Learning

• 25 March 2024, SR 35 : optional session for coding newcomers

• 22 April 2024, SR 35 : data reading and analysis with pandas

• 06 May 2024, SR 35 : simple machine learning approaches from scratch: markov chains, linear regression

• 03 June 2024, SR 35 : more machine learning from scratch - KNN & k-means clustering, q-learner

• 10 June 2024, SR 35 : neural networks from scratch & common machine learning libraries

• 17 June 2024, SR 33 : final project planning, recap & course reflection (Attention: different seminar room)

Refer to the session chapters to follow up on what we did in the specific sessions, and check out the general references section for further material.

Contents:

• Kick-off session
• Initial self-learning phase
  • Setup & Python basics
  • Loops and functions
  • Reading and analysing .csv data sets
• Pretty Plotting Pandas
  • Demo example
• Markov chains and linear regression
  • Markov chains to generate nonsense(?) texts
  • Getting into linear regression
• K-Nearest-Neighbours and K-Means Clustering
  • KNN (K-nearest-neighbour) classification
  • K-means clustering
• Neural networks
  • Basic Python Percetpron Demo
  • More realistic basic neural networks
• Exercises
  • Exercise 1 - Basic Python applied to data wrangling
  • Exercise 2 - Pretty plotting with Pandas
  • Exercise 3 - Markovian funsense or linear depression?
• References
  • Core Topic
  • Python Basics
  • Official Python Documentation
  • Tools
  • Extended Materials
• Notes on grading
  • Notes on the final project
  • Notes on nitpickyness and transparency

This documentation was produced with Sphinx, the Python Documentation Generator. Sphinx itself is written in Python and can be installed with pip install sphinx. Additionally, for this specific docs, also the MyST Parser is used in order to write the section files in Markdown. linkify-it-py is used so that URLs are automagically converted to links. For the theme I am relying on the good old sphinx_rtd_theme, which is used by many open source projects for their documentation on readthedocs.io. This is extended by sphinx_rtd_dark_mode, so that readers get the option to switch between a light and dark mode (see the round button with the moon/sun symbole in the lower right corner of the browser tab). Theme related links can be found in the footer..

The source of this documentation, including the sphinx config used to generate this output, is available under a CC-BY-SA 4.0 license on https://gitlab.com/jackieklaura/hands-on-machine-learning