Predictive accuracy is measured using various statistical metrics such as sensitivity, specificity, positive predictive value, and negative predictive value. These metrics assess the ability of a model to correctly identify true positives (e.g., toxic effects correctly predicted) and true negatives (e.g., non-toxic effects correctly predicted). Additionally, ROC curves and the area under the curve (AUC) are commonly used to evaluate the overall performance of predictive models in toxicology.
