Euclidean distance

Euclidean distance is the square root of the summed squared differences between two vectors — the straight-line distance you would measure with a ruler.

It is the default notion of “how far apart” in most of machine learning, and it is a contraction: square the differences, sum over the feature index, take the root.