A confusion matrix is a convenient solution to present the forms of mistakes a machine learning mode makes. It’s an N by N grid with numbers, where the worth within the [n, m] cell represents the variety of examples annotated with the n-th class recognized because the m-th class. On this tutorial, I’ll concentrate on making a confusion matrix and a heatmap. The colour palette can be used to display the sizes of various groups, making it easy to note similarities or significant differences in group sizes. This type of visualization is handy whenever you take care of quite a few categories.
Here’s a visual explanation of the weather of the confusion matrix.
