R

Curva ROC

ROC

A ROC curve is a graphical representation of a model's diagnostic ability across different thresholds.

Curva ROC

The Receiver Operating Characteristic (ROC) curve is a graphical tool used to evaluate the performance of a modelo de clasificación binaria. It illustrates the trade-off between sensitivity (true positive rate) and specificity (1 – tasa de falsos positivos) en varios umbrales.

Cada punto en la curva ROC representa un umbral diferente para clasificar las instancias en clases positivas y negativas. El eje x del gráfico muestra la tasa de falsos positivos (FPR), mientras que el eje y muestra la tasa de verdaderos positivos (TPR). Un modelo con clasificación perfecta alcanzaría la esquina superior izquierda del gráfico, indicando un 100% de sensibilidad y un 0% de tasa de falsos positivos.

El área bajo la curva ROC (AUC) is a key metric derived from the ROC curve. AUC values range from 0 to 1, where a score of 0.5 indicates no discriminative ability (similar to random guessing), and a score of 1.0 indicates perfect classification. Generally, a higher AUC value signifies better rendimiento del modelo.

ROC curves are particularly useful in medical diagnostics, credit scoring, and any field where binary decisions are made based on probabilistic outputs. By analyzing the ROC curve, practitioners can choose an umbral óptimo that balances sensitivity and specificity according to the specific context of their application.

oEmbed (JSON) + /