References: What machine learning actually is

Source material

Source material (conceptual spine):
• StatQuest with Josh Starmer: "A Gentle Introduction to Machine Learning"
  Creator: Josh Starmer
  YouTube: https://www.youtube.com/watch?v=Gv9_4yMHFhI
  Channel / site: https://statquest.org/
  License: as published on StatQuest's public YouTube channel (link-out only)

Source material (hands-on companion):
• Microsoft: "ML For Beginners" (Introduction module)
  Repository: https://github.com/microsoft/ML-For-Beginners
  License: MIT

Clawdemy provides original notes, summaries, and quizzes derived from this material
for educational purposes. All rights to the original videos and curriculum remain
with their creators.

What this lesson draws from each source

This lesson is a synthesis of two sources, and it is worth being precise about which idea comes from where:

• StatQuest’s “A Gentle Introduction to Machine Learning” anchors the framing that machine learning is about using data to make predictions, and it is the source of this lesson’s governing rule: a method is judged by how it performs on data it was not trained on, and a fancier method is not automatically better. That train-versus-test idea, planted here, is what Phase 4 of this track is built on.
• Microsoft’s ML-For-Beginners Introduction module anchors the supervised / unsupervised taxonomy and the “when is machine learning the right tool” framing. Its applied, scikit-learn-based lessons are the natural next step if you want to write the code, not just hold the concepts.

The “neither, just write a rule” category and the two-question classification test are Clawdemy’s own framing, added to make the decision concrete; neither source presents them in exactly that form.

Going deeper

• StatQuest with Josh Starmer. The full catalog this track mirrors, intuition-first and worked step by step. Nearly every lesson in this track maps to one of these videos; following StatQuest’s machine learning videos in parallel is the single best companion to the written track.
• Microsoft ML-For-Beginners. A free, MIT-licensed, project-based curriculum that builds the same classical algorithms in Python with scikit-learn. Where this track keeps you at the level of intuition, ML-For-Beginners is where you go to build and run the models yourself.

Adjacent topics

• Linear regression (the next lesson). Now that you can classify a problem as supervised or unsupervised, the next lesson makes the simplest supervised algorithm concrete: fitting a straight line to data.
• Reinforcement learning. The third paradigm named but not covered here. If you want a sense of it, search for an introduction to agents learning from reward; it is a large field with its own canon and sits outside classical machine learning.

Community discussion

None selected for this lesson. The public introductions to “what is machine learning” have been well absorbed into the StatQuest and Microsoft resources above; a marginal forum thread does not add durable value. If a canonical discussion surfaces, it will be added at the next review.