Build on Serverless - Continuous Learning for ML

Опубликовано 3 октября 2018, 17:40

Do you have millions of observations? That's great! But after you’ve trained a model how do you keep it up to date as new training data becomes available? How do you automate and enable your models to be continuously learning? Come help us build a data snapshot pipeline that captures and transforms data before producing an updated trained model using Amazon SageMaker. You’ll use AWS Glue to capture data from a relational database, publish it, transformed, to a private S3 bucket, and then initiate a training job to update an existing model. 

Learn More about Serverless:
aws.amazon.com/serverless

Watch past Build on Serverless broadcasts:
aws.amazon.com/developer/commu...

See the upcoming Build on Serverless Schedule:
aws.amazon.com/developer/commu...

Join the Serverless conversation on Slack:
awssamopensource.splashthat.co...

Get an overview of the Serverless Applications Lens:
d1.awsstatic.com/whitepapers/a...

Github:
github.com/aws-samples/serverl...

Step Functions Tutorial:
docs.aws.amazon.com/step-funct...

RSVP for future Build on Serverless episodes:
twitch.tv/events/umX-DTI8Q3qWl...