From the course: AWS Certified Machine Learning Engineer Associate (MLA-C01) Cert Prep

Unlock this course with a free trial

Join today to access over 24,700 courses taught by industry experts.

Hands-on learning: Amazon SageMaker Clarify

Hands-on learning: Amazon SageMaker Clarify

From the course: AWS Certified Machine Learning Engineer Associate (MLA-C01) Cert Prep

Start my 1-month free trial Buy for my team

Hands-on learning: Amazon SageMaker Clarify

“

(inspiring music) - [Instructor] Hello, guys. And welcome again. In today's lesson, we're going to talk about the Amazon SageMaker Clarify, how to use it in order to detect bias and explain predictions. What we're going to do in this lab, we're going to first train an XGBoost model using a training dataset. Then, we're going to measure the pre-training bias and the post-training bias. Remember, the pre-training bias is the bias in the dataset before model training, while the post-training bias is the bias in the model predictions after training the model. Then, we're going to analyze feature importance. So we're going to explain the predictions, and we're going to access and review the generated bias and explainability reports using the Amazon SageMaker Clarify, which generated those reports. Here, we're using a CSV data format, and an alternative method of using this notebook would be to run the SageMaker Clarify jobs using the AWS SDK for Python, but we're not going to use that in…

Contents