Contact Me!
Azure Machine Learning Studio (ML Studio) gives you the ability to create experiments to generate machine learning models on existing data.
But first, you must get data into ML Studio. ML Studio runs in the Azure cloud; so, if that data is on your local hard drive, you will need to import it.
You can do so by creating a new data source.
Sign into Machine Learning Studio and select "DATASETS" from the left menu, as shown in Fig. 1.
Fig. 1
To create a new dataset, click the [+NEW] button (Fig. 2) at the bottom left of the screen.
Fig. 2
From the popup menu, DATASET | FROM LOCAL FILE, as shown in Fig. 3.
Fig. 3
The "Upload a new dataset" dialog displays, as shown in Fig. 4.
Fig. 4
Click the [Browse…] button and select a file from your local computer and click the [Open] button, as shown in Fig. 5.
Fig. 5
This closes the "File Open" dialog and returns you to the "Upload a new dataset" dialog, as shown in Fig. 6.
Fig. 6
At the "ENTER A NAME FOR THE NEW DATASET" field, enter a name by which you wish to refer to this dataset in your ML experiments. This defaults to the filename on your computer.
At the "Select a dataset type…" dropdown, select the format of the file you selected.
Click the Check button when finished.
The file uploads to the cloud server and is listed in the "DATASETS" tab, as shown in Fig. 7.
Fig. 7
Once you have uploaded a file as a dataset, it is available within any of your experiments. From within an ML Experiment, expand "Saved Datasets" and "My Datasets". Your file should be listed under "My Datasets", as shown in Fig. 8.
Fig. 8
You can drag this dataset onto your experiment design surface to work with it.
In this article, I showed how to create a dataset, based on a file on your local computer.