Metric
Domain: Evaluation
🎯 Motivation
A specific measurement of model performance: accuracy, precision, recall, F1, latency. Different metrics for different needs.
📋 Concrete Example
Metrics for a classifier: accuracy (90%), precision (88%), recall (92%), F1 (90%). Each measures a different aspect. Choosing the right metric is crucial.
🔗 Analogy
Fitness metrics for an athlete - speed (100m time), strength (bench press), endurance (marathon), agility (shuttle run). Different metrics for different aspects. AI metrics work similarly.