Calculus for ML
Derivatives, gradients, and the chain rule — the engine behind every gradient-based learning algorithm.
5 lessons
01Derivatives and Gradients
How functions change — from single-variable derivatives to multi-dimensional gradients, the foundation of every learning algorithm.
Chain Rule and Backpropagation
How gradients flow backward through composed functions — the chain rule is the only math you need to understand backpropagation.
Multivariable Optimization
Finding minima in high-dimensional landscapes — critical points, convexity, the Hessian, and why most ML losses are non-convex.
Jacobians & Vector Calculus
Extending derivatives to vector-valued functions — the Jacobian matrix, the chain rule for layers, and how backpropagation really works.
Calculus for ML Quiz
Test your understanding of derivatives, the chain rule, and gradient-based optimization.