Retraining
The process of training a model again on updated data to restore or improve its performance. Retraining addresses model drift and incorporates new patterns the original model did not learn.
Why It Matters
Regular retraining keeps models accurate as the world changes. Without it, every model has an expiration date — the question is how quickly it degrades.
Example
Retraining a fraud detection model monthly with the latest transaction data so it recognizes new fraud patterns that have emerged since the last training.
Think of it like...
Like a doctor attending continuing education — the medical field evolves, and staying current requires ongoing learning, not just initial training.
Related Terms
Model Drift
The gradual degradation of a model's predictive performance over time as the real-world environment changes. Model drift can be caused by data drift, concept drift, or both.
Data Drift
A change in the statistical properties of the input data over time compared to the data the model was trained on. When data drifts, model predictions become less reliable.
Model Monitoring
The practice of continuously tracking an ML model's performance, predictions, and input data in production to detect degradation, drift, or anomalies after deployment.
MLOps
Machine Learning Operations — the set of practices that combine ML, DevOps, and data engineering to deploy and maintain ML models in production reliably and efficiently.