Azure Machine Learning is a cloud predictive analytics service which makes it possible to create predictive solutions based on algorithms.
How to get Dataset in Azure Machine learning
There are several sample datasets included with the ML Studio that you can use or you can import your dataset from any source into your workspace. To get dataset create a new experiment by clicking +NEW at the bottom left of window and select blank experiment.
Here experiment is the default name, so you can change it as per your requirement of the project. In the Search box type the dataset you want to include from the saved dataset list. Here I would like to add automobile price data from dataset and to view the data, right click on the output port of the dataset and select the visualize.
It will look as follows:
Prepare the Data or Clearing the values not Required
A dataset requires some pre-processing before it can be analysed, as you can see some of the values in the above window are missing, so we will clean these values to analyse the data accurately.
Steps to be followed:
- Add a module from the search box, which will remove the required column completely. For example, select columns, this module will allow you to include or exclude the columns of dataset in the model.
Connect the output port of the first dataset (Automobile price data) to the input port of the second dataset (Select columns in dataset)
Select the Second Dataset (Select columns in dataset) and on the right side, in the properties panel click Launch Column Selector
- Click with Rules on left
- Click All Columns Under Begin with
- From drop-downs select Exclude and Column Names, and click inside the text box to select the columns you want to exclude as selected here Normalized losses.
- In the properties, exclude columns from the dataset will be indicated.
- Right click on the second dataset (Select columns in dataset) and select visualize to view the data. It will show the dataset with removed column.
So, by doing modification in the dataset you can have a clear analytical view of the data.