O'Reilly logo
live online training icon Live Online training

An Introduction to Amazon Machine Learning on AWS

Best Practices for Building Machine Learning Datasets and Models to Make Predictions

Asli Bilgin

Artificial Intelligence has been around for many decades. With the advent of cloud computing, inexpensive compute and store have made it possible for machines to learn more efficiency and quickly. Amazon Machine Learning (Amazon ML) is built on the highly scalable and highly available. Amazon Web Services cloud platform. In this course, we will show you how to harness this speed and power to build effective Amazon Machine datasets and models that you will use to make predictions on real world data.

We will provide you with the tools and knowledge to apply the learnings of this course to your own scenarios. You’ll see how you can benefit from the ability of a machine to make predictions on future data through hands on labs and key concepts. Together we will walk through Amazon ML’s end to end workflow using practical and pragmatic business scenarios.

You don’t have to be a data scientists or developer to benefit from Amazon ML. Although it will be helpful to have some prior experience with AWS, it isn’t required to take this course. AWS is continuing to make great strides to innovate their Artificial Intelligence and Machine Learning Platform. The concepts learned in this course will provide you with the foundation to build your own innovative systems on this dynamic platform.

What you'll learn-and how you can apply it

  • Understand appropriate use Cases than can benefit from Amazon ML
  • Build datasets, models and fine tune them by learning tips and tricks for supervised machine learning
  • Walk away with a strong understanding of Machine Learning taxonomy and how it works behind the scenes
  • Couple the conceptual knowledge with the hands on experience to generate real time and batch predictions

This training course is for you because...

  • Beginning to to Advanced Technologists or those with a passion for technology
  • Hobbyist or professional technologists
  • Business owners and–entrepreneurs with strong domain knowledge
  • Those interested in Machine Learning on AWS – you do not need to be a developer or data scientist; however you should be knowledgeable about organizing data in digital form.

Prerequisites

  • Proficiency in programming helpful, but not needed.
  • Experience with AWS will be helpful.
  • Experience with configuring console based systems, such as the AWS Console would be very beneficial.
  • Having previous experience with technology platforms, such as cloud computing will be helpful.
  • Here are some Live Training courses you can take (search by title on O'Reilly Online Learning, for the latest course dates):

Getting Started with Amazon Web Services (AWS) (with Richard Jones or Chad Smith)

AWS Certified Cloud Practitioner Crash Course

Course Set-up

  • Must have an AWS account – there is a small fee for Machine Learning that will be charged if you chose to follow along with the labs using the AWS console.
  • Software, such as Microsoft Excel or Google Sheets, that will enable you to view CSV (comma separated value) files.

Recommended Preparation

Recommended Follow-up

  • Amazon Machine Learning LiveLessons by Asli Bilgin (available soon in O'Reilly Online Learning)
  • Review Machine Learning videos from Amazon re:Invent 2018

About your instructor

  • Asli Bilgin is an award-winning cloud computing executive who has over two decades of experience working for companies such as Dell, Microsoft and Amazon. Her firm, Nokta Consulting, specializes in IT transformation and modernization leveraging disruptive technologies such as cloud computing, machine learning and blockchain. At Amazon, Asli created, launched and ran the global Software as a Service program At Microsoft, she led the cloud and web strategy for 80 countries in the Middle East & Africa, based out of Dubai. Asli is a passionate advocate for the impact technology can make on people’s lives.. Asli was the architect behind the LEGO and Microsoft partnership effort for WomenBuild, a program to promote compute science as an art and science, specifically for girls and women.

Schedule

The timeframes are only estimates and may vary according to how the class is progressing

Day 1

Segment 1: Amazon Machine Learning Basics (1.5 hour)

  • What is Machine Learning?
  • The Amazon ML & AI Platform
  • Which Use Cases Can Amazon ML Solve?
  • How Does Amazon ML Work?
  • Curator Project Sample Business Problem for this Live Training

Break - 15 minutes

Segment 2: Prepare data for ML Processing (1.5 hours)

  • Interactive Lab: Prepare Data
  • Interactive Lab: Configure S3
  • Interactive Lab: Amazon Machine Learning Dashboard

Break - 15 minutes

Segment 3: Configure Data Source and Schema (30 min)

  • Machine Learning Taxonomy
  • Configure Data Source
  • Refine schema

Segment 4: Configure Model (30 min)

  • Create Model from Data Insights Page
  • Configure Model Settings
  • Types of ML models

Q&A - 30 mins

Day 2

Segment 1: Evaluating a Model (30 min)

  • Options for refining a model
  • What Happens in an Evaluation?

Segment 2: Interactive Lab - Evaluate the Model (30 min)

  • Amazon ML Model Summary
  • Model Insights: Evaluation Summary
  • Model Insights: Evaluation Alerts
  • Evaluate Model Performance

Break - 15 minutes

Segment 3: Predictions (30 min)

  • How do Predictions Work?
  • What are the Types of Predictions?
  • Batch Predictions
  • Real-time Predictions

Segment 4: Interactive Lab: Real Time Predictions (15 min)

  • Learning objectives
  • Create Real-time Predictions
  • View Results

Break - 15 minutes

Segment 5: Interactive Lab: Batch Predictions (1 hour)

  • Learning objectives
  • list text hereCreate Prediction Data
  • Choose Model for Prediction
  • Create Prediction Datasource
  • Choose Destination for Predictions
  • Batch Prediction Summary
  • View the Manifest
  • Download Prediction Results from S3
  • View Prediction Results
  • Apply the Prediction

Q&A - 30 minutes