Evaluating binary classification models
Model metrics tell you how accurate your models are and some of the areas they struggle in. In classification problems, many of these metrics can be embodied into a single graphic – the confusion matrix.
Confusion matrixes take the form of a table that tells you how your model classifies test data and what data points its getting confused by and incorrectly classifying.
In machine learning, we refer to our predictions as either true or false, based on their correctness, and positive or negative, based on our prediction. This is shown in the following table:
|
Predicted true |
Predicted false |
|
|
Actually true |
True positive |
False negative |
|
Actually false |
False positive |
True negative |
Table...
