The supervised setup, the cost function, and gradient descent doing the work.
8h · 11 lessons · 1 challenge
Before this: Array Programming & Vectorization, Calculus & Optimization
The first model, taken slowly. What supervised learning is, what a cost function measures, and what gradient descent does to it, one step at a time.
Then the parts that make it usable on real data: several features at once in vectorized form, feature scaling and why it changes how the run behaves, and the engineered and polynomial features that let a straight-line model fit a curve.
Ten residual plots. Name what is wrong with each model from the residuals alone, before you are shown the data.
Worth 350 XP.
