Amazon Machine Learning for Human Activity Recognition


546 students completed the lab in ~1h:12m

Total available time: 2h:0m

144 students rated this lab!

Machine Learning is a very powerful technology to drive your data-driven decisions

Lab Overview

Nowadays it's possible for anyone to exploit the huge volumes of information available through big data and open datasets, whether you are a data scientist, an enterprise developer, or a small startup. Amazon Machine Learning lets you focus completely on your data, without wasting your time with countless trial models or complicated math. Amazon Machine Learning as a service offers great potential in making applications smarter, by making it easy to use for developers of all skill levels.

This Lab will offer a very brief overview of the main machine learning concepts, and then use an open dataset from UCI to train and use a real-world model for HAR (Human Activity Recognition). You will walk through the whole process, from the dataset analysis and Datasource creation, all the way to model training/evaluation, and execute a Python script to generate real time predictions. This Lab will give you a general understanding of machine learning that should lead to ideas of how to use Amazon Machine Learning to build out models for your own environment.

Lab Objectives

Upon completion of this Lab you will be able to:

  • Understand basic Machine Learning (ML) concepts and ML phases (process flow)
  • Know the importance of data manipulation prior to using Amazon ML
  • Create a Datasource for ML input
  • Use the Amazon ML Management Console to visualize your data
  • Explore the performance of your ML model
  • Use a test ML model to perform a prediction based on new input data

Lab Prerequisites

You should be familiar with the following:

  • Connecting to an EC2 instance
  • Basic file edits and working from the Linux command-line
  • Python basics (or similar scripting language)
  • Conceptual understanding of Amazon Software Development Kit (SDK) and Application Programming Interface (API) is helpful, but not required


Lab Environment

Before completing the lab instructions the environment will look as follows:

  • A running EC2 instance
  • Note: This lab is serverless. The EC2 instance is started and configured for you by Cloud Academy simply for convenience sake. 


After completing the lab instructions the environment should look similar to:

Follow these steps to learn by building helpful cloud resources

Logging in to the Amazon Web Services Console

Your first step to start the Lab experience

Machine Learning Concepts

What is machine learning and how it can help you with data-driven decisions

Dataset selection and manipulation

How to prepare your dataset for Amazon Machine Learning

Creating a Datasource, Model, and Evaluation

Datasource creation and attributes analysis, Model creation, and run an Evaluation

Training and Evaluating your first Model

Gauging the performance of your Model's Evaluation

Generating online Predictions

Use your Model to generate Predictions based on new input data

Cleaning up the ML Environment

Learn how to keep your ML dashboard clean for future evaluations