Creating an AutoML experiment in Azure ML
AutoML can be used for training regression, classification, forecasting, or CV models. For the first three, the input dataset is expected to be tabular, but for CV, you'd work with images. To take one task as an example when exploring how to work with AutoML, let's look at forecasting.
In Chapter 4, Forecasting Time-Series Data, we already covered forecasting in Power BI. We learned that Power BI can create a forecast for time-series data based on trends and seasonality that it can find in the target value itself, the target value being the number of tourists in the Netherlands per month from 2012 until 2019. Building on that example, we'll add other features to the dataset to train a forecasting model with AutoML that also takes other information into consideration when forecasting the number of tourists.
The dataset we'll use for AutoML has four columns, as follows:
- StartMonth: The first day of the month...
