How would you measure the performance of our classification model? Yeah, I know we already talked about that: employing a confusion matrix. This is the right answer, or at least part of the right answer, since in the case of random forest classification algorithms you can also check another metric, the out-of-bag error rate (OOB error rate). We can actually find both of them directly through the print method of the function. Let's check this by simply printing out the random_forest object:

random_forest

As you can see, we have here a summary of the relevant parameters and then the two really useful pieces of information we were looking for:

• OOB estimate of error rate
• The confusion matrix