Classification metrics

Earlier in the chapter, we explored choosing the best of a few nearest neighbors instances based on the number of neighbors, n_neighbors, parameter. This is the main parameter in nearest neighbors classification: classify a point based on the label of KNN. So, for 3-nearest neighbors, classify a point based on the label of the three nearest points. Take a majority vote of the three nearest points.

The classification metric in this case was the internal metric accuracy_score, which is defined as the number of classifications that were correct divided by the total number of classifications. There are alternate metrics, and we will explore them here.

Getting ready

1. To start, load the Pima diabetes dataset from the UCI repository:

import pandas as pd

data_web_address = "https://archive.ics.uci.edu/ml/machine-learning-databases/pima-indians-diabetes/pima-indians-diabetes.data"

column_names = ['pregnancy_x',
'plasma_con',
'blood_pressure',
'skin_mm',
'insulin',
'bmi',
'pedigree_func...