I would like to create a model in ML.net with unknown count of columns.
I load the data from a text file. The first column will always be the result label. After that, it can be 2, 3 or even 20 columns with custom names(Header).
I tried creating a class like this
public class NewData
{
[LoadColumn(0)]
public float Label { get; set; }
[LoadColumn(1, 100)]
[VectorType(100)]
public float[] Features { get; set; }
}
I wrote 100 because it will never be more than 100 columns.
After that, how could I create the pipeline and the model?
Would I need to concatenate all the columns inside Features?
Thanks