Doing the Work Section 019
Baselines and Choosing a Model
Start with something dumb that works, then find out honestly whether anything fancier is actually better.
9 of 9 lessons published Three reading levels on every lesson
Lessons in order
Work top to bottom. Each lesson assumes the one above it.
- The baselines you must beat first
- Build the whole pipeline with a fake model
- Logistic regression, gradient boosting, or a neural net?
- Freeze, fine-tune, or train from scratch?
- Buying accuracy with size, and when to stop
- A comparison that actually proves something
- Reading your own errors
- Evaluating by slice, not by average
- Deciding a model is good enough