课程: Responsible AI with Amazon SageMaker AI
今天就学习课程吧!
今天就开通帐号,24,700 门业界名师课程任您挑!
Challenge: Identify bias in data - Amazon SageMaker教程
课程: Responsible AI with Amazon SageMaker AI
Challenge: Identify bias in data
(upbeat music) - It's time to apply what you've learned to your first hands-on challenge. For this challenge, you'll identify pre-training bias in your data. This means you'll catch bias long before a model is trained and placed in production. First, load the data to S3. You'll find the dataset in the 01_ 05 folder in the course's GitHub repo. Next, import the data to Data Wrangler. Don't forget to pre-process your data if needed. Then create an analysis and add the features or facets to analyze. Finally, generate a bias report to determine if bias exists in your dataset. Good luck with your first challenge. I can't wait to show you how I approach this in the next video.
随堂练习,边学边练
下载课堂讲义。学练结合,紧跟进度,轻松巩固知识。
内容
-
-
-
Understanding bias in machine learning4 分钟 48 秒
-
(已锁定)
Exploring SageMaker Clarify for bias detection4 分钟 4 秒
-
(已锁定)
Evaluating bias in pretraining data4 分钟 18 秒
-
(已锁定)
Detecting bias in post-training and production2 分钟 5 秒
-
(已锁定)
Challenge: Identify bias in data56 秒
-
(已锁定)
Solution: Identify bias in data5 分钟 31 秒
-
-
-
-