Update lecture to summer semester 2023

[melsigl]

Changes so far include, but may not limit to:

• Update year to 2023.
• Classification exercise: make buy_computer dataset consistent w/Quinlan's original dataset. This closes Issue Classification: Make Example Dataset "buys_computer" Consistent to Quinlan's Original Dataset #60, see also PR Fix Information Gain for student #55. I will close PR Fix Information Gain for student #55 once this PR is merged.
• Prologue: updated exercise and lecture dates, preliminary schedule, StudOn links

See also potential comments under this PR.

[melsigl]

@dominik-probst: Feel free to include any updates that you have from your side to this PR, or - alternatively - create a separate PR.

[dominik-probst]

Looks good. I have one small comment, which I consider so insignificant that you may decide whether to merge directly or to adjust it.

[dominik-probst]

@dominik-probst: Feel free to include any updates that you have from your side to this PR, or - alternatively - create a separate PR.

Since I still need a few days for my adjustments (and will also spread them over several weeks until the semester starts) it is probably better to start a separate PR for my adjustments.

[dominik-probst]

Since my planned SS23 changes have been finished since yesterday, I have taken the liberty of attaching them to this PR.

These changes are primarily:

• Exercises:
  • Completely new assignments for the Frequent Patterns and Clustering exercises, where students must first apply procedures by hand before being asked to implement them.
  • Restructuring the Frequent Patterns and Clustering exercises to incorporate the new tasks in a more meaningful way.
  • Introduction of a table of contents in all my exercises to make it easier to navigate through the exercises.
  • Introduction of task numbering in all of my exercises to allow easier communication with students in the exercise sessions (last summer semester some students had problems communicating which task they were asking a question about in the larger exercises, this is now easily done via "I have a question about task 15").
  • Harmonisation of the spelling of some terminology (e.g. Apriori).
  • Minor corrections/changes in several of my exercise sheets.
• Lectures:
  • Harmonisation of the spelling of some terminology (e.g. Apriori).
  • Shortening of individual lectures (especially Frequency Patterns and Clustering), with content that stands on its own being moved to the appendix.
  • Minor corrections/changes in several of my lectures.
• Maintenance:
  • We had several Dependabot alerts (#1 - #4) due to security vulnerabilities in the Python packages of our exercises. I updated the corresponding requirements.txt files and tested the exercises with the new packages and found no problems.
  • The GitHub actions also issued warnings when they were run, as some of the images were out of date. Here, too, I introduced the latest variants and verified that the actions ran correctly.

In the course of today I have also added the following changes to this PR:

• Exercises:
  • As per our Microsoft Teams conversation I have added information on the organizational details of the task sheet (optional/no extra exercise session) in the first task sheet. Also, in sync with my task sheets, I have introduced a table of contents and task numbering as well. (25b293a)
  • I also introduced the table of contents and task numbering in your classification exercise to establish some consistency. (2c0f73a)
• Lecture:
  • Delete all information on the 2.5 ECTS KDD module as this is no longer offered. I have also mapped the changes in the exercise planning. (de97472)

Since I already reviewed your changes (including yesterday's schedule change) and found them to be good, I would now ask you to review my changes @melsigl .

Since I made the introduction of the table of contents and the task numbering in your two exercises without consultation with you, I have also indicated the commit here in each case, so that you can easily undo these changes, should they not have been in your interest.

[melsigl]

Thank you, @dominik-probst, for your effort to update our lecture slides and our exercises.

Your changes sound valid to me even though I had not the luxury to invest the time it deserves to read every change in all its detail. However, I have some (minor) notes:

Exercises:

• JupyterNotebook offers numbered headings, albeit as a plugin. I suspect you already thought of using this plugin, however, it may be hard to enforce its usage. Therefore, numbering our headings manually to have consistent numbers makes sense. We just have to make sure in the future, to also update these numbered headings should we introduce changes that may affect those numbers. I guess, however, that may seldom be the case.
• I did take a look at exercise 3 (frequent paterns) and stumbled across some printing functions in 3.2-A-Closer-Look-at-A-Priori-and-FP-Growth.ipynb. More specifically, print_tree of class FPTree, print_subtree of class RootNode, print_subtree of class ItemNode, print_table of class HeaderTable, and print_element of class HeaderTableElement. I understand why these exist. You may also want to consider implementing the special method __str__ which purpose is to return a human-readable and thus, an informal representation of an object. Contrary to __str__, Python also provides __repr__, which aims to provide a string representation of an object as well, albeit with the purpose of recreating said object by employing the built-in repr() function. It seems to me that you aim for a human-readable string representation and thus __str__ may be more suitable.

Lectures:

• The lecture on frequent patterns still contains some textual references to papers that are listed in our bibliography. These could be referenced via footnotes in the future.

I also had to add a fix to the classification exercise involving Naive Bayes.

All in all, consider my consent to approve & merge as I cannot change the reviewer to myself and merge without bypassing our branch protection.

[dominik-probst]

Thank you @melsigl for the review.

I have just implemented your advice on the __str__ functions.

The suggestion regarding the footnotes will be implemented somewhere in the future (Issue #69).

After the checks have been run, I will merge this PR.