K nearest neighbours (KNN) notebook applied to the Iris dataset

The lecture notes described how KNN works and this was supported by output from a notebook that applied KNN to the Iris data.

You can find this notebook here. Note that the Iris dataset is one of the standard datasets supplied with the scikit-learn library, so it is not provided here.

This notebook uses more advanced python, including

user-defined functions (specified using the def keyword)
interactive widgets that use controls to select from a selection of plots and other outputs
more complex plotting scenarios, including colour maps, grids, marker types, etc.

Students are invited to run this notebook and to examine the outputs, paying particular attention to the use of scikit-learn functions to create, configure and fit KNN models.

For convenience, the decision boundary plots for k = 1,3,5,7,17 are shared here.