AutoML Featurizer handling of label column

## Problem

Currently when using the `Featurizer` with `ColumnInformation` from `ColumnInferenceResults` there is no easy way to exclude the label column from featurization or automatically featurize it as well. Therefore you either need to pass in an IDataView with the dropped columns or you need to individually provide the column type parameters (i.e. `numericColumns`).

```csharp
var pipeline =
ctx.Auto().Featurizer(
data: dataView,
numericColumns: columnInferenceResults.ColumnInformation.NumericColumnNames.ToArray(),
outputColumnName: "Features",
excludeColumns: new[] { "Label" })
.Append(ctx.Auto().MultiClassification());

```

## Proposal

1. Automatically featurize label column (this is often the case for classification scenarios where values need to be converted to `KeyType`).

2. Create an overload of `Featurizer` with the following signature:

```csharp
Featurizer(IDataView data, ColumnInformation columnInformation, string[] excludeColumns, string outputColumnName)
```