Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 8

Amazon

Sage
Maker
AWS MACHINE LEARNING STACK
Involves 3 steps
• Building
• Training
• Hosting

AMAZON SAGE
MAKER
COMPONENTS
Use Building
Sagemaker's
hosted
Notebook Apachespark
instance or
SageMaker
Spark SDK

Console for
a point and AMAZON SAGE
click Your own
experience device(EC2 ,
laptop)
MAKER
COMPONENTS
Training
AMAZON SAGE MAKER
COMPONENTS
Hosting can be done by Built in algorithms
• XG Boost, FM linear and forecasting for supervised learning
• Kmeans , PCA for clustering and pre processing
• Image classification with convolutional networks

TensorFlow and ApacheMXNet Docker Container


• Sample data
• explore and refine models in single Notebook Instance
Hosting
• Use the same code to train on full dataset in a cluster of instances
• deploy to production AMAZON SAGE MAKER
COMPONENTS
Bring your own algoritm:
• Pick the preferred network
• Add algoritm code to a docker container
• publish to ECS
1. Start with a sample notebook

STEPS INVOLVED
2. Modify to access data sources
3. Train your model
4. Deploy your model

You might also like