5. The Pipeline
• Raw data -> prepped data -> loaded data
• Data -> data source -> model
• Model ->
– Evaluation
– Batch prediction
– Real-time prediction
7. AML Use Cases
• Easy intro to ML
• Exploration of ML and data
• Structured text data
– Especially if already in AWS
• Limited configurability
• No infrastructure
• Inexpensive
8. Alternatives
• AWS
– Image recognition
– Voice recognition
– Spark
– Deep learning AMI
• Non AWS
– Google, Azure, startups
– Open source (scikit-learn)
9. Questions?
• Thanks
– UCI for the data
• And where to find me
– dan@mooreds.com
– @mooreds
– http://www.mooreds.com/wordpress/
– AML O’Reilly video course:
https://bitly.com/introtoaml
– https://github.com/mooreds/amazonmachinelearning
-anintroduction/