Source: wikibot/detection-error-tradeoff

= Detection error tradeoff
{wiki=Detection_error_tradeoff}

The Detection Error Tradeoff (DET) curve is a graphical representation used in the fields of signal detection theory, machine learning, and statistical classification to visualize the trade-offs between various types of errors in a binary classification system. It helps to understand the performance of a classifier or detection system in varying conditions. The DET curve plots two types of error rates on a graph: 1. **False Negative Rate (FNR)**: This is the probability of incorrectly classifying a positive instance as negative.