Predict whether or not a flight will be delayed based on airports and weather. Can be built and run in a guest account.
Written by a Microsoft employee. When you are helping a group get acquainted with Azure Machine Learning, it can be useful to lead them through building an experiment. There is some functionality that isn't available in Guest Accounts (such as the Execute R Script module), and this can make some of the samples in the Gallery unbuildable. This particular example was designed to be compatible with the limitations of a Guest Account. This experiment is divided up into a number of discrete chunks. These are captured in three intermediate checkpoints and a start point that let students catch up if the fall behind, join late, or have technical difficulties. ![Experiment Overview] **Start Point** (checkpoint 0) This loads a specially formatted version of the flight delay data set available in Azure ML. **Checkpoint 1** This path cleans and organizes the flight delay data. ![Checkpoint 1] **Checkpoint 2** This path cleans and organizes the weather data. !![Checkpoint 2] **Checkpoint 3** This part merges the flight and weather data. !![Checkpoint 3] ** Full solution** Finally, the machine learning models are created, trained and evaluated. ![Machine learning] : https://github.com/brohrer-ms/public-hosting/raw/master/flight_delays/image1.png : https://github.com/brohrer-ms/public-hosting/raw/master/flight_delays/image2.png : https://github.com/brohrer-ms/public-hosting/raw/master/flight_delays/image3.png : https://github.com/brohrer-ms/public-hosting/raw/master/flight_delays/image4.png : https://github.com/brohrer-ms/public-hosting/raw/master/flight_delays/image5.png