About this resource
This page provides context on the authors, and the values guiding the structure, tone, and design of this resource.
1 Land Acknowledgement
The authors of this text are machine learning teaching and research faculty at the University of Toronto. The text is thus written on the land on which the University of Toronto operates. For thousands of years it has been the traditional land of the Huron-Wendat, the Seneca, and the Mississaugas of the Credit. Today, this meeting place is still the home to many Indigenous people from across Turtle Island. Many of the authors are first- and second-generation immigrants, and we are still learning about the history of this land and its people. As such, in writing this text, we aimed to use data sets that are collected on this land, and showcase the stories of its people (some of this is currently asperational in this current draft). We are grateful to have the opportunity to live and work on this land.
3 Acknowledgements
We gratefully acknowledge the following individuals for their valuable feedback on draft versions of this resource:
- Prof. Randy Hickey
- Prof. Dan Zingaro
4 Generative AI Usage
For some chapters, GenAI models were used to assist in the writing process. In particular, much of the code used in the visualizations were created with GenAI assistance. The human authors have vetted content produced by GenAI, and take responsibility for the content.
5 What We Valued
While there are many machine learning texts out there, we wanted a resource that is approachable for learners who need some help remembering the math, but still goes into the depth necessary to fully understand key concepts. We wanted a resource that leveraged what interactive web technologies can offer, in order to help develop geometric intuition. We were inspired by resources like distill.pub and other interactive explanations of ML content.
The three main values below guided many of our design decisions.
5.1 Value: Accessibility
Accessible and Rigorous: The set of course notes (web textbook) is intended to be accessible to students of various mathematical preparation and abilities, without compromising the depth and rigor of the content.
Visually Rich but Accessible: The notes should use interactive visualizations to help explain and demonstrate key concepts. Care should be taken to only include visualizations and interactions to demonstrate ideas. Colors should be chosen to be color-blind friendly, and items should be distinguished not only by color, but also by other features (e.g., shape). Avoid using italics. Bold text should be used for definition only, not emphasis.
Accessibility Features: All figures should be captioned. Although the visualization and interaction should be rich, the text should also read well on a screen reader. Alt text should be provided for images and diagrams.
5.2 Value: Reduce Cognitive Load
Purpose: All interaction, math, definitions, etc. should have a pedagogical purpose. Avoid default interactions like zoom/pan, mouseover, etc.
Skim-able: Section headings should be meaningful. Paragraphs should be meaningful, with topic sentences where appropriate.
Dataset: Where possible, the same dataset should be used, unless there is a pedagogical reason for introducing a new data set.
Cleanliness: Where possible, reduce lines and borders.
Boxes: Only two kinds of boxes for emphasis: “Question” and “Definition”. Avoid other kinds of boxes.
Length: Each “chapter” of the text should be cohesive, but not too long.
5.3 Value: Encourage Active Reading and Choice
Question: Where appropriate, include questions for students to test their understanding before continuing. These should be straightforward questions, rather than long active learning exercises.
Interactivity: The interactivity should provide students with an option to develop further intuition about the materials.
[Stretch Goal] Math & Definitions: Explain mathematical concepts, ideally in a way where students can interactively prompt for additional explanation.