ML 6.2Lesson outlineLesson 2 of 3 in Module 6
Naive Bayes
Classify by multiplying per-feature likelihoods under an independence assumption we know is false, and make it work in practice.
Video in production.
Lesson outline: Naive Bayes
In this lesson12 parts
- 01Spam from word counts
- 02The counts per class
- 03What to beat
- 04Bayes' theorem, one word at a time
- 05A prior and a likelihood per word
- 06Maximum likelihood is a share of the counts
- 07Training is counting, by hand
- 08Predicting in logs
- 09Measuring it
- 10Where the assumption shows
- 11Code, and the sealed messages
- 12Recap, your challenge, and what comes next
Builds on
This lesson stands on its own. These go deeper into what it uses.
- Stats 3.2–3.3: Conditional Probability and Independence
- 4.2
- 4.3
- 5.2
- Maths 1.2: Exponents and Logarithms
- 5.1
- ML 4.5: Hyperparameter Search
- 5.2–5.3
- 6.1 (Stats 4.3 moved ahead of this video in the learner path
- 2026-10-17
- for Gaussian Naive Bayes)