Recordar, also known as sensitivity or true positive rate, is a key performance metric used to evaluate the effectiveness of a classification model in the campo de inteligência artificial and aprendizado de máquina. It quantifies the ability of a model to correctly identify all relevant instances within a dataset.
Matematicamente, recall é definido como a razão entre os resultados de verdadeiro positivo e o número total de instâncias positivas reais. A fórmula para recall é:
Recall = Verdadeiros Positivos / (Verdadeiros Positivos + Falsos Negativos)
Nesta fórmula:
- Verdadeiros Positivos (TP) referem-se às instâncias que são corretamente classificadas como positivas.
- Falsos Negativos (FN) are the instances that are incorrectly classified as negative, despite being positive.
Recall is particularly important in scenarios where the cost of missing a positive instance is high, such as in medical diagnoses or detecção de fraudes. A high recall score indicates that the model is effective at capturing most of the relevant instances, while a low score suggests that many positive instances are being overlooked.
However, it is essential to consider recall in conjunction with other metrics, such as precision (the ratio of true positives to the total predicted positives) and F1 score (the média harmônica of precision and recall), to get a comprehensive understanding of a model’s performance. Balancing recall and precision is crucial, as focusing solely on maximizing recall may lead to a high number of false positives.