Supported Metrics

Beyond volume and speed, model evaluation metrics provide insights into predictive ability, generalization capability, and overall input-output quality. The evaluation metrics you choose to rely upon heavily depend on your model's specific problem domain, type of data, and desired outcome.

Although TruEra supports the most commonly used ML metrics, the platform is designed to focus on identifying the influences that have the most impact on model performance and reliability. Put another way, the features in a data collection directly influence the predictive performance of the models using the data. Understanding these influences aids in explaining a model's predictions.

The list of platform-supported metrics with definitions includes:

• Performance
• Fairness
• Drift

Click Next below to continue.