These notes walk through the calculus toolkit that underpins machine learning optimization:
- Derivatives & Partial Derivatives: Single-variable and multivariate differentiation, geometric intuition of slopes and tangent planes.
- Chain Rule: Composing derivatives through nested functions — the backbone of backpropagation.
- Gradient Descent: Computing gradients, learning rate intuition, and iterative parameter updates.
- Multivariate Optimization: Hessians, saddle points, convexity, and conditions for minima.