From the course: Machine Learning with SageMaker by Pearson
Unlock this course with a free trial
Join today to access over 25,200 courses taught by industry experts.
Model A/B testing demonstration - Amazon SageMaker Tutorial
From the course: Machine Learning with SageMaker by Pearson
Model A/B testing demonstration
As we saw in the lesson, AB testing is deploying two different versions of a model to an endpoint and then distributing traffic across those two models in order to compare and contrast, and then choose the best model that you're going to put into production. I have a demo here. It's somewhat halfway complete in that the outcome of actually doing the analysis of the data is a little bit too deep and involved for the requirements of this course. So I'll explain when I get to that point. But what we're going to do here, following along with the demo guide here, we're going to take the models that were created as part of a previous lesson, and as the underfitting and overfitting demos. I have three models that were created as part of that demo. And what I'm going to do is I'm going those three to a SageMaker endpoint. You'll notice I said those three to a single endpoint. And what that does is it will actually launch the three models on the target hardware specified. And then based on a…
Contents
-
-
-
-
-
-
-
(Locked)
Learning objectives39s
-
(Locked)
Model evaluation metrics: Accuracy, precision, and recall9m 39s
-
(Locked)
Using SageMaker Clarify for bias detection and interpretability7m 40s
-
(Locked)
Comparing model performance using A/B testing5m 47s
-
(Locked)
Model A/B testing demonstration6m 26s
-
(Locked)
Managing model versions with SageMaker Model Registry5m 55s
-
(Locked)
Model registry demonstration10m 48s
-
(Locked)
-
-
-
-
-