Creating the datasource

When working with Amazon ML, the data always resides in S3 and it is not duplicated in Amazon ML. A datasource is the metadata that indicates the location of the input data allowing Amazon ML to access it. Creating a datasource also generates descriptive statistics related to the data and a schema with information on the nature of the variables. Basically, the datasource gives Amazon ML all the information it requires to be able to train a model. The following are the steps you need to follow to create a datasource:

1. Go to Amazon Machine Learning: https://console.aws.amazon.com/machinelearning/home.
2. Click on getting started, you will be given a choice between accessing the Dashboard and Standard setup. This time choose the standard setup:

Perform the following steps, as shown in the following screenshot:

1. Choose an S3 location.
2. Start typing the name of the bucket in the s3 location field, and the list folders and files should show up.
3. Select the titanic_train.csv file.

 

4. Give...