Forecast Flight Delays with Amazon Machine Learning


92 students completed the lab in ~36m

Total available time: 1h:0m

37 students rated this lab!

Lab Overview

Amazon Machine Learning provides an easy-to-use interface for creating certain types of machine learning models. This Lab uses Amazon Machine Learning to create a regression model. Regression models predict a continuous variable, as opposed to a set of classes that are predicted by classification models. You will use US Department of Transportation flight data to forecast flight delays with the Amazon Machine Learning regression model. You will create two models using custom model recipes: one that can be used for long-term forecasting and one for near-term forecasting.

Lab Objectives

Upon completion of this Lab you will be able to:

  • Create and analyze the performance of regression models in Amazon Machine Learning
  • Define custom schemas and recipes for Amazon Machine Learning datasources and models
  • Make real-time predictions with Amazon Machine Learning models

Lab Prerequisites

You should be familiar with:

  • Basic S3 concepts
  • Some knowledge of machine learning concepts is beneficial, but not required

Lab Environment

Before completing the Lab instructions, the environment will look as follows:

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

Understanding the Flight Delay Data

Gain an understanding of the data you will use in this Lab

Creating an Amazon Machine Learning Flight Datasource

Create a datasource using the flight data and schema

Creating a Pair of Flight Delay Machine Learning Models

Create long- and near-term arrival delay prediction models

Inspecting the Flight Delay Datasource

See statistics and distributions for the datasource attributes

Evaluating the Flight Delay Models

Evaluate and compare the two flight delay models

Forecasting Flight Delays with the Models

Use the two models to forecast delays and understand trends