Confusion matrix

It is the visualization of a predictive classification model that provides answers to two straightforward questions, each with a positive or negative classification.

To what extent did the model accurately predict reality?

• True Positive (TP) model correctly predicted the positive scenario of an event.

For instance, the model predicted a transaction to be fraudulent, which was subsequently confirmed as fraud by the customer.

• True Negative?(TN) model correctly predicted the negative scenario of an event.

For instance, the model predicted a transaction was not fraudulent, and the customer did not report it as fraud.

To what extent did the model inaccurately predict reality??

• False Positive (FP) model wrongly predicted the positive scenario of an event.

For instance, the model predicted a transaction to be fraudulent, which the customer did not report as fraud.

• False Negative?(FN) model wrongly predicted the negative scenario of an event.

For instance, the model predicted a transaction as legitimate, but the customer reported it as fraud.

By now, I'm sure you have a sense of why it's called a “Confusion Matrix”