Model Evaluation Patterns

What problem does it solve?

This skill provides a unified framework to quantify and compare model performance across classification, regression, and survival tasks using yardstick and related R packages, enabling consistent benchmarking and reproducible reporting.

Core Features & Use Cases

• Classification metrics: accuracy, AUC, F1, precision/recall, and confusion matrices for binary and multiclass problems.
• Regression metrics: RMSE, MAE, R², MAPE, and calibration tools for regression predictions.
• Visualization & benchmarking: ROC/PR curves, calibration plots, and side-by-side model comparisons for rapid decision making.

Quick Start

Evaluate a trained model by computing standard metrics and generating a performance report across relevant tasks.