Getting Started with Machine Learning on AWS

Machine Learning (ML) is an exciting field that Cloud Computing has helped to accelerate. AWS has played a big part in this with it’s continually expanding range of services from the simply named Machine Learning through to SageMaker. But how do you get started? Thankfully you don’t need to become an expert in linear algebra or statistics, all you need to begin is good idea of the life-cycle of a ML project and a passing familiarity with these AWS services. In this talk we’ll outline a typical ML project and review services such as SageMaker and Rekognition so that you can begin to make use of them in your own projects.

Transcript

1. @mlfowler_ @Claranet Getting Started with Machine Learning Mike Fowler - Senior Site Reliability Engineer - Public Cloud Practice PLACE CUSTOMER LOGO HERE
2. London PostgreSQL Meetup January 24th 2019 About Me About Me
3. @mlfowler_ @Claranet
4. @mlfowler_ @Claranet • What is Machine Learning? • The AWS Machine Learning Stack • ML Use Cases • Machine Learning: The Forgotten Service Agenda
5. @mlfowler_ @Claranet Ethics Source: https://peakcare.wordpress.com/2011/10/05/heads-in-the-sand/
6. @mlfowler_ @Claranet What is … Machine Learning?
7. @mlfowler_ @Claranet How do Machines Learn? Source: https://towardsdatascience.com/machine-learning-types-2-c1291d4f04b1
8. @mlfowler_ @Claranet Machine Learning Concepts • Models - Mathematical equation with a solution space approximating the outputs for the given inputs • Feature Engineering - Process of identifying & creating features from the data that will influence/assist the model • Training - Repeated process attempting to find the model that is “just right” such that it does not overfit or underfit the training data
9. @mlfowler_ @Claranet Dang it Jim, I’m an Engineer not a Mathematician!
10. @mlfowler_ @Claranet The AWS Machine Learning Stack
11. @mlfowler_ @Claranet Categories of ML Services
12. @mlfowler_ @Claranet Categories of ML: Ready to Eat Amazon Comprehend Amazon Polly Amazon Rekognition Amazon Textract Amazon Transcribe Amazon Translate
13. @mlfowler_ @Claranet Use Case: Audio Description for Images
14. @mlfowler_ @Claranet Use Case: Audio Description for Images Lambda
15. @mlfowler_ @Claranet Use Case: Audio Description for Images Rekognition Image Lambda
16. @mlfowler_ @Claranet Use Case: Audio Description for Images Polly Rekognition Image Lambda
17. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles
18. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles S3
19. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles S3 Lambda
20. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles S3 Rekognition Text in Image Lambda
21. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles
22. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles
23. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles
24. @mlfowler_ @Claranet Use Case: Solving Sudoku Puzzles Have a Play! http://bit.ly/comsumsudoku https://github.com/gh-mlfowler/sudokusolve
25. @mlfowler_ @Claranet Categories of ML: Part Baked Amazon Forecast Amazon Personalize
26. @mlfowler_ @Claranet The Forgotten Service
27. @mlfowler_ @Claranet Scene: Being On Call https://www.silicon.co.uk/wp-content/uploads/2017/02/Pager.jpg
28. @mlfowler_ @Claranet Scene: Our Engineer Rests Peacefully Source: https://peakcare.wordpress.com/2011/10/05/heads-in-the-sand/ https://i.pinimg.com/originals/cb/32/5f/cb325f9c268bf2135125f512d95
29. @mlfowler_ @Claranet Engineer Resting Scene: Red Alert! Source: https://peakcare.wordpress.com/2011/10/05/heads-in-the-sand/ https://vignette.wikia.nocookie.net/memoryalpha/images/6/6b/RedAlert.jpg/revision/latest?cb=20100117050244&path-prefix=en
30. @mlfowler_ @Claranet Scene: Peace Source: https://peakcare.wordpress.com/2011/10/05/heads-in-the-sand/ https://www.lakelouiseinn.com/wp-content/uploads/2019/01/LakeLouise2-1.jpg
31. @mlfowler_ @Claranet
32. @mlfowler_ @Claranet Identify a Problem to Solve Many PagerDuty incidents resolve before I respond disrupting my sleep needlessly
33. @mlfowler_ @Claranet Source Relevant Data
34. @mlfowler_ @Claranet Input Data
35. @mlfowler_ @Claranet Input Data
36. @mlfowler_ @Claranet Target
37. @mlfowler_ @Claranet Target
38. @mlfowler_ @Claranet Target
39. @mlfowler_ @Claranet Train the Model
40. @mlfowler_ @Claranet The Lambda Architecture Master Data Serving Layer Batch Layer Speed Layer S3 EMR Kinesis Streams Glue Redshift (Batched Views) DynamoDB (Real-Time Views)
41. @mlfowler_ @Claranet ML Model The Lambda Architecture + ML Master Data Serving Layer Batch Layer Speed Layer S3 EMR Kinesis Streams Glue Redshift (Batched Views) DynamoDB (Real-Time Views) Amazon Machine Learning
42. @mlfowler_ @Claranet Make Predictions
43. @mlfowler_ @Claranet Make Predictions
44. @mlfowler_ @Claranet Make Predictions
45. @mlfowler_ @Claranet Make Predictions
46. @mlfowler_ @Claranet Categories of ML: Raw Ingredients Amazon SageMaker Amazon Elastic Inference Amazon SageMaker Ground Truth AWS Deep Learning AMIs Apache MXNet on AWS TensorFlow on AWS
47. @mlfowler_ @Claranet Feature Engineering • Most models only take numeric input • Values often need to be constrained - Scale Min/Max - Logarithm • Some values can’t be used - Identifiers - Attributes that wouldn’t be known when making a prediction
48. @mlfowler_ @Claranet Fin
49. @mlfowler_ @Claranet Questions ? Have a Play! http://bit.ly/comsumsudoku https://github.com/gh-mlfowler/sudokusolve