vimeo.com/.../463429576

A k-nearest neighbor model
is a modeling approach

where we take an observation
and find closely related

other observations
or neighbors,

and use information about
those neighbors to make a prediction.

For instance,

if a given flower is very similar
to five other flowers,

that are classified
as flower type A,

it would stand to reason
we can predict our given flower

is also flower type A.