Optimization Basics (Gradient Descent Intuition) — Roadmap Step & Res…
The core idea of iteratively improving a model by following the gradient downhill
- Level: beginner
- Estimated time: 1 days
- Roadmap: Machine Learning
Before this step
Interview questions
- Why can a learning rate that's too large actually prevent a model from converging?
Study resources
- Gradient Descent, Step-by-Step (Video) — StatQuest's widely-loved, step-by-step explanation of gradient descent.
- Convex Optimization Overview (PDF) — Stanford CS229's reference notes on the optimization concepts underlying ML.
Part of
- Mathematics & Statistics for ML — section
- Machine Learning roadmap — the full learning path