Neural Network Projects with Python

Confusion matrix

The confusion matrix is a useful visualization tool that provides analysis on the true negative, false positive, false negative, and true positives made by our model. Beyond a simple accuracy metric, we should also look at the confusion matrix to understand the performance of the model.

The definition of true negative, false positive, false negative, and true positives are as follows:

• True negative: Actual class is negative (no diabetes), and the model predicted negative (no diabetes)
• False positive: Actual class is negative (no diabetes), but the model predicted positive (diabetes)
• False negative: Actual class is positive (diabetes), but the model predicted negative (no diabetes)
• True positive: Actual class is positive ...