JAX

Every machine learning model revolves around minimizing a cost function, and that minimization is pure calculus. This article covers the core calculus concepts behind ML, with hands-on code in JAX so you can see each idea in action. JAX is a hardware-accelerated numerical computing library with built-in automatic differentiation. It happens to be a great fit for exploring calculus in an ML context. Derivatives: Rate of Change A derivative measures the rate of change of one quantity with respect to another. It’s how we describe change in the real world. ...