From Confusion Matrix to Weighted Cross Entropy

Confusion matrix is a super convenient way to summarize the classification result of an ML model. As shown below, it comprises 4 sections of TP (True Positives), FP (False Positives), FN (False Negatives), and TN (True Negatives). For instance, FN in the case of COVID-19 would be the number of people with COVID-19 who were diagnosed to have no COVID-19. Here I summarize few keywords derived from confusion matrix, which appear a lot in ML papers with classification tasks. Try not to memorize, because the names themselves make a lot of sense!