Retraining a model

Retraining the model with the latest datasets will help to build a more accurate model over time. Typically, continuous monitoring of the deployed model performance will give an opportunity for retraining the model. Automating the process of retraining the model and continuous deployment of better models for real-time predictions with better predictions saves millions of dollars. In the retraining step, typically it goes through the same phase of training, testing, and validation of the model, but the sophisticated retraining process identifies the outliers to produce a better model from the previous history of the model's performance.

Get Industrial Internet Application Development now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.