Classical models built from the math you already know
So far the routine has always been the same two steps: fit some parameters to the training data, then use those fitted parameters to predict. k-Nearest-Neighbours regression skips the first step entirely. To predict a value at a new point, it looks up the k training points closest to it and simply averages their outputs. There is no formula to fit and no parameters learned in advance: all the work happens the moment a prediction is asked for.
🔒 This is a Pro lesson — the interactive figure, worked examples, quiz and practice open with Pro access.
▶ k-Nearest-Neighbours Regression